Not in the top 1,000 by GitHub stars
Prism-Shadow/GDPevo
A Benchmark for Evaluating Agent Self-Evolution on Real Business Tasks
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Local scripts
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
evaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Local scripts
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
evaluation/eval_workspace/codex_skill_creator_comparison/creators/naive/upstream/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
evaluation/eval_workspace/codex_skill_creator_comparison/creators/opencode/upstream/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Local scripts
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Local scripts
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Network or broad shell
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_01/SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
56d60ae4ae5e → 5082a7773791
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Not checked yet
First observation → 56d60ae4ae5e
No earlier version was observed for comparison.
121GitHub stars
3stars in 7 days
1.7kskills in this repository
2commits in 30 days
Heat
3.8- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
- skill-creator
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream - skill-creator
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream - skill-creator
evaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream - skill-creator
evaluation/eval_workspace/codex_skill_creator_comparison/creators/naive/upstream - opencode-skill-creator
evaluation/eval_workspace/codex_skill_creator_comparison/creators/opencode/upstream - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_013/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_01 - northstar-payer-ops
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_02 - northstar-payer-review
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_02 - asteria-fleet-data-quality-hub
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_01 - self_attempt_02
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_02 - self_attempt_03
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_02 - medbridge-sales-ops-reconcile
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02 - scn003-crm-support-ops
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02 - reflect-3-crm-support-console
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_02 - scn003-crm-support-console
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02 - apexcloud-retention-ops
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01 - reflect-3-attempt-02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01 - reflect-3
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_010/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_03 - northwind-people-lifecycle-closeout
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_01 - peopleops-lifecycle-clearance
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_02 - hr-lifecycle-clearance-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/self/self_attempt_03 - cedar-ridge-intake-verification
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/fewshot/fewshot_attempt_01 - cedar-ridge-intake-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/fewshot/fewshot_attempt_02 - cedar-ridge-intake-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/fewshot/fewshot_attempt_03 - cedar-ridge-intake-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_02 - cedar-ridge-intake-coordination
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/reflect-3/reflect-3_attempt_03 - cedar-ridge-intake-coordination
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/self/self_attempt_01 - cedar-ridge-intake-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/self/self_attempt_02 - cedar-ridge-intake-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/self/self_attempt_03 - northstar-payer-ops-determination
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/fewshot/fewshot_attempt_01 - northstar-payer-ops-determination
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/fewshot/fewshot_attempt_02 - northstar-payer-determination
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/fewshot/fewshot_attempt_03 - northstar-payer-ops
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_01 - northstar-payer-determinations
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_02 - northstar-payer-determinations
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/reflect-3/reflect-3_attempt_03 - northstar-payer-ops
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_01 - northstar-payer-operations
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_02 - northstar-payer-ops
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_03 - ehr-quality-governance-packets
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/fewshot/fewshot_attempt_01 - ehr-governance-packet
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/fewshot/fewshot_attempt_02 - ehr-governance-packet
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/fewshot/fewshot_attempt_03 - ehr-packet-generation
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_02 - ehr-referral-quality-packets
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/reflect-3/reflect-3_attempt_03 - ehr-quality-governance-packet
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/self/self_attempt_01 - ehr-quality-governance-packets
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/self/self_attempt_02 - ehr-quality-governance-packet
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/fewshot/fewshot_attempt_01 - clinic-protocol-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/fewshot/fewshot_attempt_03 - synthetic-clinic-protocol-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_01 - clinic-protocol-decision-support
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_02 - clinic-protocol-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/reflect-3/reflect-3_attempt_03 - synthetic-clinic-protocol-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/self/self_attempt_01 - clinic-runtime-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/self/self_attempt_02 - synthetic-clinic-protocol-cds
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/fewshot/fewshot_attempt_01 - investigation-review-hub-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/fewshot/fewshot_attempt_02 - investigation-review-hub-gap-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/fewshot/fewshot_attempt_03 - investigation-review-hub-gap-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01 - investigation-review-hub-gap-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02 - investigation-review-hub-gap-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_01 - investigation-review-hub-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_02 - investigation-review-hub-answer
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03 - court-operations-closeout
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/self/self_attempt_01 - court-operations-closeout
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/self/self_attempt_03 - licensing-board-structured-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/fewshot/fewshot_attempt_03 - licensing-board-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_01 - licensing-board-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02 - licensing-batch-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/self/self_attempt_03 - ma-deal-workbench-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/fewshot/fewshot_attempt_01 - ma-deal-workbench-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/fewshot/fewshot_attempt_02 - ma-deal-workbench-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/fewshot/fewshot_attempt_03 - ma-deal-workbench-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01 - ma-deal-workbench-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02 - ma-deal-workbench-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03 - ma-deal-workbench
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/self/self_attempt_01 - ma-deal-workbench-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/self/self_attempt_02 - ma-deal-workbench-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/self/self_attempt_03 - asteria-fleet-dq-certification
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/fewshot/fewshot_attempt_01 - asteria-fleet-dq-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/fewshot/fewshot_attempt_02 - asteria-fleet-dq-certification
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/fewshot/fewshot_attempt_03 - asteria-fleet-data-quality-reconciliation
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01 - asteria-fleet-data-quality-certification
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03 - asteria-fleet-dq-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/self/self_attempt_01 - asteria-fleet-dq-certification
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/self/self_attempt_02 - asteria-fleet-dq-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/self/self_attempt_03 - atlas-ops-scorecard
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_03 - atlas-commerce-ops-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_01 - atlas-commerce-ops-scorecard
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_02 - atlas-commerce-ops-scorecard
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/self/self_attempt_02 - self_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/self/self_attempt_03 - pho-algorithmic-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01 - pho-algorithmic-transport-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_02 - pho-algorithmic-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_03 - pho-transportability-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03 - pho-registered-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/self/self_attempt_01 - pho-registered-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/self/self_attempt_02 - pho-algorithmic-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/self/self_attempt_03 - portfolio-env-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01 - work-portfolio-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/fewshot/fewshot_attempt_02 - engineering-portfolio-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/fewshot/fewshot_attempt_03 - portfolio-work-review
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_02 - portfolio-engineering-audit
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03 - portfolio-env-analytics
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/self/self_attempt_01 - self_attempt_02
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/self/self_attempt_02 - portfolio-work-item-analysis
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_008/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/self/self_attempt_03 - fewshot_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/fewshot/fewshot_attempt_01 - fewshot_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/fewshot/fewshot_attempt_02 - fewshot_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/self/self_attempt_01 - self_attempt_02
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/self/self_attempt_02 - self_attempt_03
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/self/self_attempt_03 - harborcrm-front-of-funnel
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_01 - harborcrm-front-of-funnel
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_02 - harborcrm-front-of-funnel
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_03 - harborcrm-front-of-funnel-handoff
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01 - harborcrm-front-of-funnel
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02 - harborcrm-front-of-funnel
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03 - harborcrm-front-of-funnel-handoff
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/self/self_attempt_01 - harborcrm-front-of-funnel-handoff
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/self/self_attempt_02 - harborcrm-front-of-funnel-handoff
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/self/self_attempt_03 - medbridge-sales-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_01 - medbridge-sales-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_02 - medbridge-sales-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/self/self_attempt_03 - crm-support-ticket-resolution
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_01 - crm-support-ticket-resolution
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_02 - crm-support-ticket-resolution
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_03 - apexcloud-retention-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_01 - apexcloud-retention-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_02 - apexcloud-retention-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/self/self_attempt_03 - erp-finance-close-tasks
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_01 - erp-finance-close-api
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_02 - erp-finance-close-reviewer
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_03 - procureops-erp-answers
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_01 - procureops-control-answers
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02 - procureops-erp-answers
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/self/self_attempt_03 - northwind-erp-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_01 - northwind-erp-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_02 - northwind-erp-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/self/self_attempt_03 - private-wealth-advisory-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_01 - private-wealth-advisory
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_02 - private-wealth-advisory-planner
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_03 - private-wealth-advisory-structured-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01 - private-wealth-advisory-structured-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02 - private-wealth-advisory-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03 - private-wealth-advisory-planner
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_01 - private-wealth-advisory-structured-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_02 - private-wealth-advisory-planning
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_03 - crescent-finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_01 - crescent-finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_02 - crescent-finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_03 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_01 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_02 - finance-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_03 - asteria-investment-office
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_01 - asteria-investment-office-tasks
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_02 - asteria-investment-office-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_03 - asteria-investment-office-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01 - asteria-investment-office-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02 - asteria-investment-office-json-desk
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03 - asteria-investment-office-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/self/self_attempt_01 - asteria-investment-office-json-desk
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/self/self_attempt_02 - asteria-investment-office-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/self/self_attempt_03 - credit-committee-packets
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_01 - credit-committee-packets
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_02 - credit-committee-packets
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/self/self_attempt_01 - self_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/self/self_attempt_02 - self_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/self/self_attempt_03 - peopleops-console-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_01 - peopleops-console-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_02 - peopleops-console-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_03 - peopleops-console-lifecycle-control
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01 - peopleops-console-lifecycle-control
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02 - peopleops-console-case-resolution
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03 - peopleops-console-resolution
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/self/self_attempt_01 - peopleops-console-resolver
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/self/self_attempt_02 - peopleops-console-erp-hr
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/self/self_attempt_03 - cedar-ridge-intake
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_01 - cedar-ridge-intake
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_02 - cedar-ridge-intake
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_03 - intake-coordination-portal-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_01 - intake-coordination-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_02 - cedar-ridge-intake-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_03 - cedar-ridge-intake-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_01 - cedar-ridge-intake-coordination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_02 - cedar-ridge-intake-coordination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_03 - northstar-payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_01 - northstar-payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_02 - northstar-payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_03 - payer-ops-case-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_01 - payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_02 - payer-ops-structured-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_03 - northstar-payer-ops
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/self/self_attempt_01 - northstar-payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/self/self_attempt_02 - northstar-payer-ops-determination
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/self/self_attempt_03 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_01 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_02 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_03 - ehr-governance-packets
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_01 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_02 - ehr-governance-packets
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_03 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/self/self_attempt_01 - ehr-governance-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/self/self_attempt_02 - ehr-governance-packet-builder
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/self/self_attempt_03 - clinic-protocol-decision-support
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_01 - synthetic-clinic-protocol-answer
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_02 - clinic-protocol-cds
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_03 - clinic-protocol-answer
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_03_invalid_20260725T215503Z - clinic-protocol-json-assessment
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_01 - clinic-protocol-structured-response
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_02 - clinic-protocol-decision-support
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_03 - harborview-clinic-protocol-answers
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/self/self_attempt_01 - clinic-protocol-decision-support
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/self/self_attempt_02 - clinic-protocol-decision-support
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/self/self_attempt_03 - investigation-review-hub-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_01 - investigation-review-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_02 - investigation-review-gap-dashboard
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_03 - investigation-hub-production-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01 - investigation-review-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02 - investigation-review-hub-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03 - investigation-review-hub-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/self/self_attempt_01 - investigation-review-hub-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/self/self_attempt_02 - investigation-review-hub-gap-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/self/self_attempt_03 - court-ops-closeout
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_01 - court-closeout-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_02 - court-closeout-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_03 - court-closeout-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01 - court-closeout-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_02 - court-clerk-closeout-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03 - court-operations-closeout
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/self/self_attempt_01 - court-operations-closeout-packet
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/self/self_attempt_02 - court-closeout-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/self/self_attempt_03 - licensing-review-batch
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_01 - licensing-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_02 - licensing-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_03 - licensing-board-structured-decisions
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_01 - licensing-review-decisions
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02 - licensing-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03 - licensing-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/self/self_attempt_01 - licensing-review-decisions
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/self/self_attempt_02 - licensing-review-decisions
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/self/self_attempt_03 - ma-deal-workbench-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_01 - ma-deal-workbench-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02 - ma-deal-workbench-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_03 - ma-deal-workbench
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01 - ma-workbench-json-deliverables
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02 - ma-deal-workbench-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03 - ma-deal-workbench-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_01 - ma-deal-workbench-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_02 - ma-deal-workbench-json
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_03 - asteria-fleet-dq-certification
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_01 - asteria-fleet-data-quality
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_02 - asteria-fleet-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_03 - asteria-fleet-data-quality-hub
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01 - fleet-data-quality-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02 - asteria-fleet-dq-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03 - asteria-fleet-data-quality-hub
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/self/self_attempt_01 - asteria-fleet-dq-certification
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/self/self_attempt_02 - asteria-fleet-dq-reconciliation
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/self/self_attempt_03 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_01 - atlas-ops-analytics
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_02 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_03 - atlas-ops-analytical-scorecard
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_01 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_02 - ops-request-scorecard
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_03 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_01 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_02 - atlas-ops-reporting
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_03 - pho-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_01 - pho-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_02 - pho-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_03 - pho-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01 - pho-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02 - pho-registered-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03 - pho-registered-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_01 - pho-registered-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_02 - pho-registered-algorithmic-audit
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_03 - portfolio-mix-sla-release-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_01 - portfolio-mix-sla-release-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_01/skill - portfolio-env-review
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_02 - portfolio-env-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03 - portfolio-env-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/self/self_attempt_01 - portfolio-env-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/self/self_attempt_02 - portfolio-environment-analysis
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/self/self_attempt_03 - task-group-001-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-001-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02 - task-group-001-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03 - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03 - harborcrm-handoff-solver
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/self/self_attempt_02 - harborcrm-json-handoff
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_001/self/self_attempt_03 - task-group-002-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill - task-group-002-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill - task-group-002-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/self/self_attempt_02 - medbridge-sales-ops-sop
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/self/self_attempt_03 - task-group-003-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_01/support-console-demonstration - task-group-003-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_02/task-group-003-support-console-io - task-group-003-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_03 - crm-service-ticket-sop
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/self/self_attempt_01 - crm-service-ticket-sop
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/self/self_attempt_02 - crm-support-ticket-sop
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_003/self/self_attempt_03 - task-group-004-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-004-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_02/apexcloud-retention-analytics - task-group-004-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_03/apexcloud-retention-analytics - apexcloud-retention-ops-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01 - apexcloud-retention-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_02 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03 - apexcloud-retention-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_01 - apexcloud-retention-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_02 - apexcloud-retention-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_03 - task-group-005-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-005-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02 - task-group-005-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03 - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03 - erp-finance-control-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/self/self_attempt_01 - erp-finance-control-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/self/self_attempt_02 - task-group-005-erp-finance-control
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/self/self_attempt_03 - task-group-006-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-006-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02 - task-group-006-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03 - reflect-3-procureops-controls
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_01 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02 - reflect-3-procureops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_03 - procureops-json-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/self/self_attempt_01 - procureops-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/self/self_attempt_02 - procureops-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/self/self_attempt_03 - task-group-007-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-007-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02 - task-group-007-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03 - northwind-erp-operations-control
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01 - northwind-erp-structured-decisions
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02 - northwind-erp-control-desk
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03 - northwind-erp-control-desk
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/self/self_attempt_01 - northwind-erp-control-desk
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/self/self_attempt_02 - northwind-erp-decision-sop
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/self/self_attempt_03 - task-group-008-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01 - task-group-008-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02 - task-group-008-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03 - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_03 - private-wealth-advisory-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/self/self_attempt_02 - self_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_008/self/self_attempt_03 - task-group-009-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_01 - task-group-009-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_02 - task-group-009-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_03 - reflect-3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03 - crescent-finance-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/self/self_attempt_02 - crescent-finance-ops-reporting
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_009/self/self_attempt_03 - task-group-010-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_01/task-group-010-input-output - task-group-010-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_02/asteria-portfolio-risk-json - task-group-010-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_03/task-group-010-asteria-portfolio-risk - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/self/self_attempt_01 - asteria-portfolio-risk
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/self/self_attempt_02 - self_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/self/self_attempt_03 - task-group-011-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_01/input-output-credit-json - task-group-011-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_02/credit-committee-json - task-group-011-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_03/credit-office-json - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/self/self_attempt_02 - self_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_011/self/self_attempt_03 - task-group-012-fewshot-attempt-01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_01 - task-group-012-fewshot-attempt-02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_02 - task-group-012-fewshot-attempt-03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_03 - peopleops-console-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01 - peopleops-lifecycle-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02 - peopleops-console-verification
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03 - self_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/self/self_attempt_01 - self_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/self/self_attempt_02 - peopleops-lifecycle-control-solver
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/self/self_attempt_03 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_01 - cedar-ridge-intake-coordination
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_02 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_03 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_01 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_02 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/reflect-3/reflect-3_attempt_03 - cedar-ridge-intake-coordination
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/self/self_attempt_01 - cedar-ridge-intake-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/self/self_attempt_02 - cedar-ridge-intake-auditor
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/self/self_attempt_03 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_01 - skill
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_03 - payer-ops-structured-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_01 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_02 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/reflect-3/reflect-3_attempt_03 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/self/self_attempt_01 - skill
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/self/self_attempt_02 - northstar-payer-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_014/self/self_attempt_03 - ehr-quality-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_01 - ehr-quality-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_02 - ehr-quality-json-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_03 - ehr-quality-packet-builder
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_01 - ehr-quality-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_02 - ehr-quality-packet-audit
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/reflect-3/reflect-3_attempt_03 - ehr-quality-governance-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/self/self_attempt_01 - ehr-governance-json-packets
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/self/self_attempt_02 - ehr-quality-packet-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/self/self_attempt_03 - clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_01 - clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_03 - clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_01 - synthetic-clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_02 - clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/reflect-3/reflect-3_attempt_03 - clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/self/self_attempt_01 - synthetic-clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/self/self_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_016/self/self_attempt_03 - investigation-review-hub
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_01 - investigation-review-hub-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_02 - investigation-review-hub-remediation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_03 - investigation-review-hub-remediation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01 - investigation-review-gap-dashboard
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02 - investigation-review-remediation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03 - investigation-review-hub-analysis
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/self/self_attempt_01 - investigation-review-hub-remediation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/self/self_attempt_02 - investigation-review-hub-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/self/self_attempt_03 - reconcile-court-closeout
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_01 - court-packet-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_02 - court-packet-assembler
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_03 - court-closeout-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01 - court-closeout-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_02 - court-packet-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03 - court-closeout-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/self/self_attempt_01 - court-packet-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/self/self_attempt_02 - court-packet-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_018/self/self_attempt_03 - licensing-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_01 - licensing-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_02 - licensing-review-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_03 - licensing-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_01 - licensing-record-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02 - licensing-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03 - licensing-structured-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/self/self_attempt_01 - licensing-record-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/self/self_attempt_02 - licensing-review-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/self/self_attempt_03 - ma-deal-workbench
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_01 - deal-workbench-ma-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02 - deal-workbench-ma-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_03 - ma-deal-workbench-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01 - ma-deal-workbench-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02 - deal-workbench-ma-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03 - ma-deal-workbench-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/self/self_attempt_01 - ma-deal-workbench-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/self/self_attempt_02 - deal-workbench-ma-json
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/self/self_attempt_03 - asteria-hub-reconciler
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_01 - asteria-dq-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_02 - asteria-data-quality-audit
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_03 - asteria-quality-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01 - fleet-data-quality-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02 - asteria-fleet-data-quality
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03 - asteria-data-quality-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/self/self_attempt_01 - asteria-data-quality-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/self/self_attempt_02 - asteria-data-quality-reconciliation
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/self/self_attempt_03 - atlas-commerce-operations
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_01 - atlas-commerce-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_02 - atlas-commerce-ops-analytics
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_03 - atlas-ops-analytics
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_01 - atlas-commerce-ops-analysis
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_02 - atlas-ops-analytics
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/reflect-3/reflect-3_attempt_03 - atlas-commerce-ops
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/self/self_attempt_01 - atlas-ops-workplace
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/self/self_attempt_02 - atlas-ops-request-analysis
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/self/self_attempt_03 - fewshot_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_01 - pho-algorithmic-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_02 - pho-observatory-algorithmic-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_03 - reflect-3_attempt_01
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01 - reflect-3_attempt_02
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02 - reflect-3_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03 - public-health-observatory-algorithmic-audit
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/self/self_attempt_01 - public-health-observatory-audit
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/self/self_attempt_02 - self_attempt_03
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/self/self_attempt_03 - portfolio-environment-analysis
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_01 - engineering-portfolio-analysis
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_02 - portfolio-env-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_03 - portfolio-work-auditor
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_01 - engineering-portfolio-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_02 - engineering-portfolio-auditor
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03 - portfolio-environment-review
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/self/self_attempt_01 - engineering-portfolio-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/self/self_attempt_02 - engineering-portfolio-audits
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/self/self_attempt_03 - harbor-crm
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_01 - harbor-crm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_02 - harbor-crm
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_03 - harbor-crm
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_01 - harbor-crm
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_02 - harborcrm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_03 - harbor-crm-integration
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_01 - harbor-crm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_02 - harborcrm-task-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_03 - harborcrm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_01 - harbor-crm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_02 - harbor-crm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_03 - harbor-crm-integration
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_01 - harbor-crm-handoff
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_02 - harbor-crm-integration
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_02 - medbridge-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_03 - medbridge-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_03 - support-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_01 - telecom-support-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_02 - support-console-resolver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_03 - support-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_01 - telco-support-triage
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_02 - support-console-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_03 - support-console-operator
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_01 - support-console-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_02 - support-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_03 - support-console-resolver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_01 - support-console-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_02 - support-console-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_03 - support-console-resolver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_01 - support-console-resolver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_02 - support-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_02 - apexcloud-retention
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03 - finance-close-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_01 - erp-finance-close
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_02 - erp-finance-api-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_03 - finance-close-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_01 - erp-finance-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_02 - erp-finance-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_03 - erp-finance-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_01 - finance-close-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_02 - finance-close-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_03 - finance-erp-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_01 - erp-finance-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_02 - finance-operations-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_03 - finance-api-batch-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_01 - erp-finance-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_02 - finance-audit-api
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_03 - procureops-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_01 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_02 - procureops-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_03 - procureops-finops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_01 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_02 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_01 - procurops-agent
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_02 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_03 - procureops-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_01 - procureops-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_02 - procureops-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_03 - procureops-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_01 - procureops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_02 - procureops-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_03 - northwind-erp-desk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_01 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_02 - northwind-erp-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_03 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_01 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_02 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_03 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_01 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_02 - northwind-erp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_03 - northwind-erp-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_01 - northwind-erp-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_02 - northwind-erp-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_03 - northwind-erp-desk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_01 - northwind-erp-desk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_02 - northwind-erp-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_03 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_01 - wealth-advisory-planner
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_02 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_03 - wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_01 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_02 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_03 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_01 - wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_02 - wealth-advisory-planner
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_03 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_01 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_02 - private-wealth-advisory
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_03 - advisory-planner
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_01 - wealth-advisory-planner
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_02 - wealth-advisory-planner
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_03 - asteria-portfolio-analytics
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01 - asteria-investment-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_02 - asteria-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_03 - asteria-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01 - asteria-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_02 - asteria-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03 - asteria-portfolio-risk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_01 - asteria-portfolio-risk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02 - asteria-investment-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_03 - asteria-investment-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_01 - asteria-investment-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02 - asteria-portfolio-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_03 - asteria-investment-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_01 - asteria-portfolio-analytics
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02 - asteria-portfolio-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03 - credit-risk-committee-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_01 - credit-office-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_02 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_03 - credit-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_01 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_02 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_03 - credit-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_01 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_02 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_03 - credit-risk-analysis
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_01 - credit-office-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_02 - credit-office
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_03 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_01 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_02 - credit-risk-committee
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_02 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_03 - people-ops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_02 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_03 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_02 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_03 - peopleops-workflow
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_02 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_03 - peopleops-lifecycle-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_02 - peopleops-hrms-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_02 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_02 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_02 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_02 - intake-coordinator
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_02 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_03 - northstar-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_01 - northstar-payer-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_03 - northstar-payer-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_01 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_02 - northstar-payer-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_03 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_01 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_03 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_01 - northstar-payer-operations
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_03 - northstar-payer-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_01 - payer-ops-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_03 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_01 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_02 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_03 - ehr-quality-packet
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_01 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_02 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_03 - ehr-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_01 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_02 - ehr-governance-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_01 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_02 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_03 - ehr-quality-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_01 - ehr-quality-packet
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_02 - ehr-governance
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_03 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_01 - clinic-protocol-assessor
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_02 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_03 - synthetic-clinic-protocol
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_01 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_02 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_03 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_01 - clinic-protocol
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_02 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_03 - harborview-clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_01 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_02 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_03 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_01 - clinic-protocol-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_02 - clinic-decision-support
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_03 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_01 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_02 - investigation-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_03 - investigation-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_01 - investigation-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_02 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_03 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_01 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_03 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_01 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_02 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_03 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_01 - investigation-review-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_02 - investigation-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_03 - court-clerk-closeout
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_01 - court-clerk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_02 - court-clerk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_03 - court-clerk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_01 - court-clerk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_02 - court-clerk-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_03 - court-clerk-closeout
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_01 - court-clerk-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_02 - court-clerk-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_03 - court-ops-portal
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_01 - court-closeout-reconciliation
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_02 - court-clerk-closeout
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_03 - court-clerk-closeout
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_01 - court-clerk
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_02 - court-clerk-closeout
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_03 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_02 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_03 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_02 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_03 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_02 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_03 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_02 - licensing-reviewer
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_03 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_02 - licensing-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_03 - deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_01 - deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_02 - manda-deal-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_03 - ma-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_01 - ma-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_02 - ma-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_03 - ma-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_01 - ma-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_02 - m-a-deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_03 - deal-workbench-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_01 - deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_02 - deal-workbench-analyzer
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_03 - deal-workbench-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_01 - ma-deal-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_02 - deal-workbench
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_03 - asteria-fleet-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_01 - asteria-fleet-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_02 - asteria-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_03 - asteria-dqh
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_01 - asteria-fleet-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_02 - asteria-fleet-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_03 - asteria-data-quality-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_01 - asteria-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_02 - asteria-fleet-data-quality-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_03 - asteria-data-quality-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_01 - asteria-fleet-data-quality-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_02 - asteria-fleet-data-quality
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_03 - asteria-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_01 - asteria-dq-hub-certification
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_02 - asteria-dq-hub
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_03 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01 - atlas-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_02 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_03 - atlas-ops-queries
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_01 - atlas-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_02 - atlas-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_03 - atlas-ops-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_01 - atlas-ops-analytics
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02 - atlas-ops-scoring
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_03 - atlas-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_01 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_02 - atlas-operations-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_03 - atlas-ops-api
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01 - atlas-ops-api
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02 - atlas-ops-analyst
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_03 - pho-algorithmic-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01 - pho-audit-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_02 - pho-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_03 - pho-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_01 - pho-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02 - pho-audit-protocols
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03 - pho-audit-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_01 - pho-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_02 - pho-audit-portal
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_03 - pho-audit-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_01 - pho-algorithmic-transport-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_02 - pho-algorithmic-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03 - pho-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_01 - pho-audit-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_02 - pho-algorithmic-audit
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_03 - work-intelligence
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_01 - eng-portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_02 - engineering-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_03 - eng-portfolio-metrics
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_01 - eng-portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_02 - portfolio-mgmt
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_03 - eng-portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_01 - portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_02 - engineering-portfolio
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_03 - portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_01 - portfolio-review
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_02 - portfolio-management-solver
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_03 - portfolio-management
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_01 - portfolio-analyzer
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_02 - portfolio-analytics
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_03 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_01 - harborcrm-workflows
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_02 - harborcrm-data-workflows
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_03 - harborcrm-json-handoff
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_01 - harborcrm-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_02 - harborcrm-prepare
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_03 - harborcrm-structured-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_01 - harborcrm-reconcile
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_02 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_03 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_01 - harborcrm-handoff-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_02 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_03 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_01 - harborcrm-handoff
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_02 - harborcrm-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_01 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_02 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_01 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_02 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_03 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_01 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_02 - medbridge-sales-ops-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_03 - medbridge-sales-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_01 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_02 - medbridge-sales-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_03 - support-console-resolver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_01 - support-console-json-resolver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_02 - support-console-operations
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_03 - support-console-batch-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_01 - support-console
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_02 - support-console-resolver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_03 - support-console-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_01 - support-console-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_02 - support-console-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_03 - support-console-operations
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_01 - support-console-resolution
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_02 - support-console-operations-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_03 - support-console-resolver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_01 - support-console
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_02 - support-console-resolution
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_01 - apex-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_02 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_03 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_01 - apexcloud-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_02 - apex-retention-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03 - task-group-005-finance-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_01 - erp-finance-control-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_02 - task-group-005-finance-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_03 - finance-api-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_01 - erp-finance-risk-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_02 - finance-api-close-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_03 - finance-control-api-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_01 - finance-api-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_02 - erp-finance-control-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_03 - task-group-005-finance-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_01 - erp-finance-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_02 - api-finance-close-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/naive/fewshot_attempt_03 - erp-batch-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_01 - erp-finance-api-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_02 - erp-finance-release-control
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_03 - procureops-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_01 - procureops-casework
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_02 - procureops-json-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_03 - procureops-task-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_01 - procureops-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_02 - procureops-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03 - procureops-control-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_01 - procureops-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_02 - procureops-packet-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_03 - solve-procureops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_01 - procureops-control-files
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_02 - procureops-reconcile
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_03 - procureops-control-file-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_01 - procureops-json-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_02 - procureops-packet-builder
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_03 - northwind-erp-decisions
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_01 - northwind-erp-decision-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_02 - northwind-erp
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_03 - northwind-erp-desk
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_01 - northwind-erp-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_02 - northwind-erp-decisions
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_03 - northwind-erp-desk
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_01 - northwind-erp-operations
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_02 - northwind-erp-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_03 - northwind-erp-operations
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_01 - northwind-erp-control-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_02 - northwind-erp-json-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_03 - northwind-erp-control
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_01 - northwind-erp-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_02 - northwind-erp
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_03 - private-wealth-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_01 - wealth-advisory-planning
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_02 - advisory-planning-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_03 - wealth-advisory-planning
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_01 - advisory-planning-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_02 - advisory-planning-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_03 - private-wealth-advisory-planner
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_01 - private-wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_02 - private-wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_03 - private-wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_01 - wealth-advisory-api-planning
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_02 - wealth-advisory-planning-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_03 - wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_01 - wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_02 - wealth-advisory-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_03 - crescent-finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_01 - crescent-finance-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_03 - crescent-finance-ops-reports
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_01 - finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_02 - finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_01 - crescent-finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_02 - crescent-finance-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_03 - crescent-finance-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_01 - crescent-finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_02 - crescent-finance-ops-reporter
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_03 - skill
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_01 - finance-ops-json-reporter
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_02 - crescent-finance-ops-reporting
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_03 - asteria-investment-office-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01 - asteria-portfolio-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_02 - asteria-portfolio-decision-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_03 - asteria-portfolio-decisions
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01 - skill
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_02 - asteria-investment-office-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03 - asteria-investment-office-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_01 - asteria-investment-office
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02 - asteria-investment-office
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_03 - asteria-investment-json-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_01 - solve-asteria-investment-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02 - asteria-investment-office-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_03 - asteria-portfolio-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_01 - asteria-portfolio-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02 - asteria-json-workflows
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03 - credit-office-api-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_01 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_02 - credit-office-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_03 - credit-office-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_01 - credit-office-packet-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_02 - credit-office-api-pack
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_03 - credit-office-committee-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_01 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_02 - credit-office-committee
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_03 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_01 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_02 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_03 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_01 - skill
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_02 - credit-office-committee-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03 - peopleops-console-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_01 - peopleops-console
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_02 - peopleops-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_03 - peopleops-console-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_01 - peopleops-reconcile
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_02 - peopleops-console-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_03 - peopleops-console-case-workflow
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_01 - peopleops-console-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_02 - peopleops-console-controls
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_03 - peopleops-lifecycle-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_01 - peopleops-evidence-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_02 - peopleops-reconcile
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_03 - peopleops-console-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_01 - peopleops-evidence-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_02 - peopleops-evidence-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03 - cedar-ridge-intake-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_01 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_02 - cedar-ridge-intake-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_01 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_02 - cedar-ridge-intake-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_03 - cedar-ridge-intake
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01 - cedar-ridge-intake-coordination
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_02 - cedar-intake-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_03 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_01 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_02 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_03 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_01 - cedar-ridge-intake-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_02 - cedar-ridge-intake-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_03 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_01 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_03 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_01 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_03 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_01 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_02 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_03 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_01 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_02 - northstar-payer-json-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_03 - northstar-payer-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_01 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_02 - northstar-payer-ops-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_03 - ehr-quality-packet-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_01 - ehr-quality-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_02 - ehr-quality-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_03 - ehr-quality-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_01 - ehr-quality-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_02 - ehr-quality-governance
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_03 - ehr-quality-json-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_01 - ehr-quality-governance
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_02 - ehr-quality-packet
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03 - ehr-quality-packet-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_01 - ehr-quality-json-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_02 - ehr-quality-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_03 - ehr-quality-json-packets
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_01 - ehr-quality-packet
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_02 - ehr-quality-governance
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_03 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_01 - clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_02 - synthetic-clinic-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_03 - clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_01 - clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_03 - clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_01 - synthetic-clinic-cds
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_03 - synthetic-clinic-json-protocol
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_01 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_03 - clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_01 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_02 - synthetic-clinic-protocol-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_03 - investigation-review-hub-remediation-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_01 - investigation-review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_02 - investigation-review-hub-gap-analysis
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_03 - investigation-review-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_01 - investigation-review-hub
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_02 - review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_03 - review-hub-remediation-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_01 - investigation-review-hub-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02 - investigation-review-hub
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_03 - investigation-review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_01 - investigation-review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_02 - investigation-review-hub-remediation-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_03 - investigation-review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_01 - review-hub-legal-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_02 - investigation-review-hub-remediation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_03 - court-closeout-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_01 - court-operations-closeout
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_02 - court-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_03 - court-portal-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_01 - court-closeout-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_02 - court-closeout-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_03 - court-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_01 - court-closeout-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_02 - court-ops-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_03 - court-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_01 - court-closeout-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_02 - court-closeout-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/naive/fewshot_attempt_03 - court-closeout-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_01 - court-closeout-reconciler
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_02 - court-closeout-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_03 - licensing-data-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_01 - licensing-env-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_02 - licensing-json-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_03 - licensing-review-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_01 - licensing-review-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_02 - licensing-review-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_03 - licensing-json-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_01 - licensing-record-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_02 - licensing-record-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_03 - licensing-review-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_01 - licensing-review-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_02 - licensing-record-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_03 - licensing-record-json-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_01 - licensing-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_02 - licensing-review-decisions
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_03 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_01 - ma-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_02 - ma-deal-workbench
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_03 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_01 - ma-deal-workbench-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_02 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_03 - ma-deal-workbench
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_01 - ma-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_02 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_03 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_01 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_02 - ma-deal-workbench-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_03 - deal-workbench-ma-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_01 - mna-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_02 - ma-deal-workbench-json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_03 - asteria-fleet-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_01 - asteria-fleet-dq-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_02 - asteria-hub-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_03 - asteria-hub-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_01 - asteria-dq-hub-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_02 - asteria-fleet-dq
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_03 - asteria-fleet-data-quality
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_01 - asteria-dq-auditor
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_02 - asteria-data-quality-hub
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_03 - asteria-fleet-dq-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_01 - asteria-hub-quality-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_02 - asteria-hub-reconcile
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_03 - asteria-hub-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_01 - asteria-hub-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_02 - asteria-fleet-reconciliation
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_03 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01 - atlas-ops-sql-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_02 - atlas-commerce-ops-sql
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_03 - atlas-commerce-ops-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_01 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_02 - atlas-commerce-ops-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_03 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_01 - atlas-commerce-ops-analysis
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_03 - atlas-commerce-ops-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_01 - atlas-commerce-ops-sql-analyst
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_02 - atlas-ops-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_03 - atlas-commerce-ops
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01 - atlas-commerce-workplace
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02 - atlas-ops-json-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_03 - pho-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01 - public-health-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_02 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_03 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_01 - pho-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02 - pho-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03 - public-health-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_01 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_02 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_03 - pho-portal-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_01 - solve-pho-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_02 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03 - public-health-observatory-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_01 - pho-registered-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_02 - pho-algorithmic-audits
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_03 - engineering-portfolio-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_01 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_02 - engineering-portfolio-auditor
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_03 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_01 - engineering-portfolio-audit
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_02 - engineering-portfolio-review
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_03 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_01 - portfolio-environment-analyst
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_02 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_03 - engineering-portfolio-environment
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_01 - engineering-portfolio-json-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_02 - portfolio-environment-analysis
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_03 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_01 - engineering-portfolio-env-solver
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_02 - engineering-portfolio-env
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_03
How it works
skill-creator
When your agent uses it
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
What it loads into context
Frontmatter
68 tokensAlways available to the agent
SKILL.md
7,241 tokensWhen the skill is activated
Reference documents
9,476 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Communicating with the user
- Creating a skill
- Running and evaluating test cases
- Improving the skill
- Advanced: Blind comparison
- Description Optimization
- Claude.ai-specific instructions
- Cowork-Specific Instructions
- Reference files
Scripts executed separately
eval-viewer/generate_review.pyscripts/__init__.pyscripts/aggregate_benchmark.pyscripts/generate_report.pyscripts/improve_description.pyscripts/package_skill.pyscripts/quick_validate.pyscripts/run_eval.pyscripts/run_loop.pyscripts/utils.py
Files
LICENSE.txtdoc11,357 bytesSKILL.mdcore33,168 bytesSKILL.mdcore33,168 bytesSKILL.mdcore33,168 bytesSKILL.mdcore33,168 bytesSKILL.mdcore33,168 bytesSKILL.mdcore33,168 bytesagents/analyzer.mddoc10,376 bytesagents/comparator.mddoc7,287 bytesagents/grader.mddoc9,049 bytesassets/eval_review.htmlasset7,058 byteseval-viewer/generate_review.pyruns16,365 byteseval-viewer/viewer.htmlasset44,998 bytesreferences/schemas.mddoc12,061 bytesscripts/__init__.pyruns0 bytesscripts/aggregate_benchmark.pyruns14,386 bytesscripts/generate_report.pyruns12,847 bytesscripts/improve_description.pyruns11,116 bytesscripts/package_skill.pyruns4,234 bytesscripts/quick_validate.pyruns3,972 bytesscripts/run_eval.pyruns11,464 bytesscripts/run_loop.pyruns13,605 bytesscripts/utils.pyruns1,661 bytes
Read SKILL.md
# Skill Creator
A skill for creating new skills and iteratively improving them.
At a high level, the process of creating a skill goes like this:
- Decide what you want the skill to do and roughly how it should do it
- Write a draft of the skill
- Create a few test prompts and run claude-with-access-to-the-skill on them
- Help the user evaluate the results both qualitatively and quantitatively
- While the runs happen in the background, draft some quantitative evals if there aren't any (if there are some, you can either use as is or modify if you feel something needs to change about them). Then explain them to the user (or if they already existed, explain the ones that already exist)
- Use the `eval-viewer/generate_review.py` script to show the user the results for them to look at, and also let them look at the quantitative metrics
- Rewrite the skill based on feedback from the user's evaluation of the results (and also if there are any glaring flaws that become apparent from the quantitative benchmarks)
- Repeat until you're satisfiedskill-creator
When your agent uses it
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
What it loads into context
Frontmatter
61 tokensAlways available to the agent
SKILL.md
3,921 tokensWhen the skill is activated
Reference documents
494 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- About Skills
- Core Principles
- Skill Creation Process
Scripts executed separately
scripts/generate_openai_yaml.pyscripts/init_skill.pyscripts/quick_validate.py
Files
LICENSE.txtdoc11,358 bytesSKILL.mdcore18,664 bytesagents/openai.yamlasset192 bytesreferences/openai_yaml.mddoc2,130 bytesscripts/generate_openai_yaml.pyruns6,614 bytesscripts/init_skill.pyruns14,483 bytesscripts/quick_validate.pyruns3,293 bytes
Read SKILL.md
# Skill Creator
This skill provides guidance for creating effective skills.
## About Skills
Skills are modular, self-contained folders that extend Codex's capabilities by providing
specialized knowledge, workflows, and tools. Think of them as "onboarding guides" for specific
domains or tasks—they transform Codex from a general-purpose agent into a specialized agent
equipped with procedural knowledge that no model can fully possess.
### What Skills Provide
1. Specialized workflows - Multi-step procedures for specific domainsskill-creator
When your agent uses it
Guide for creating effective skills that extend agent capabilities with specialized knowledge, workflows, or tool integrations. Use this skill when the user asks to: (1) create a new skill, (2) make a skill, (3) build a skill, (4) set up a skill, (5) initialize a skill, (6) scaffold a skill, (7) update or modify an existing skill, (8) validate a skill, (9) learn about skill structure, (10) understand how skills work, or (11) get guidance on skill design patterns. Trigger on phrases like "create a skill", "new skill", "make a skill", "skill for X", "how do I create a skill", or "help me build a skill".
What it loads into context
Frontmatter
177 tokensAlways available to the agent
SKILL.md
4,126 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core Principles
- Skill Creation Process
Scripts executed separately
scripts/init_skill.pyscripts/quick_validate.py
Files
SKILL.mdcore19,277 bytesSKILL.mdcore19,277 bytesSKILL.mdcore19,277 bytesscripts/init_skill.pyruns13,104 bytesscripts/quick_validate.pyruns5,026 bytes
Read SKILL.md
# Skill Creator
### Skill Location for Deepagents
The deepagents CLI loads skills from five sources, listed here from lowest to highest precedence:
| # | Directory | Scope | Notes |
|---|-----------|-------|-------|
| 0 | `<package>/built_in_skills/` | Built-in | Ships with deepagents CLI |
| 1 | `~/.deepagents/<agent>/skills/` | User (deepagents alias) | Default for `deepagents skills create` |
| 2 | `~/.agents/skills/` | User | Shared across agent tools |
| 3 | `.deepagents/skills/` | Project (deepagents alias) | Default for `deepagents skills create --project` |
| 4 | `.agents/skills/` | Project | Shared across agent tools |
skill-creator
When your agent uses it
Create a reusable skill from staged training examples and matching answers.
What it loads into context
Frontmatter
20 tokensAlways available to the agent
SKILL.md
71 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore356 bytes
Read SKILL.md
Use the staged training examples and their matching answers to create a reusable skill for solving similar tasks. Write it to `/work/skill/SKILL.md` with YAML frontmatter containing `name` and `description`. Follow `creator_contract.md`.opencode-skill-creator
When your agent uses it
Create, test, evaluate, optimize, and package OpenCode skills with the opencode-skill-creator plugin. Use when users explicitly mention opencode-skill-creator, OpenCode Skill Creator, creating an OpenCode skill, editing an OpenCode SKILL.md, running skill evals, benchmarking skill performance, or optimizing an OpenCode skill description. Do not use for generic Claude Code or Superpowers skill creation unless the user asks to port that workflow to OpenCode.
What it loads into context
Frontmatter
106 tokensAlways available to the agent
SKILL.md
7,093 tokensWhen the skill is activated
Reference documents
9,482 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Communicating with the user
- Creating a skill
- Running and evaluating test cases
- Improving the skill
- Advanced: Blind comparison
- Description Optimization
- Skill Installation
- Available plugin tools
- Reference files
Files
SKILL.mdcore32,472 bytesagents/analyzer.mddoc10,376 bytesagents/comparator.mddoc7,287 bytesagents/grader.mddoc9,103 bytesreferences/schemas.mddoc12,057 bytestemplates/eval-review.htmlasset7,058 bytes
Read SKILL.md
# OpenCode Skill Creator
A skill for creating new skills and iteratively improving them.
At a high level, the process of creating a skill goes like this:
- Decide what you want the skill to do and roughly how it should do it
- Write a draft of the skill
- Create a few test prompts and run opencode-with-access-to-the-skill on them
- Help the user evaluate the results both qualitatively and quantitatively
- While the runs happen in the background, draft some quantitative evals if there aren't any (if there are some, you can either use as is or modify if you feel something needs to change about them). Then explain them to the user (or if they already existed, explain the ones that already exist)
- Use the `skill_serve_review` tool to show the user the results for them to look at, and also let them look at the quantitative metrics
- Rewrite the skill based on feedback from the user's evaluation of the results (and also if there are any glaring flaws that become apparent from the quantitative benchmarks)
- Repeat until you're satisfiedWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,602 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- Normalization Rules
- Sponsor Status Derivation
- Lead Qualification (Event Tasks)
- Trade Show Prospecting
- Batch Import Cleaning
- Badge-Level Handling (Event Reconciliation with Badges)
- Sorting Rules (by task type)
- General Rules
Files
SKILL.mdcore15,435 bytesSKILL.mdcore15,435 bytesSKILL.mdcore15,435 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
Base URL: `http://34.46.77.124:8001`. All API calls use this prefix. Never start local env/setup.sh or use localhost — environment_access.md overrides any task text that mentions local URLs.
## API Reference
### Events (post-event reconciliation tasks)
| Endpoint | Returns |
|---|---|
| `GET /api/events` | All events |
| `GET /api/events/{event_id}` | Single event with `end_date`, `followup_days_after_end`, `sponsor_followup_days_after_end`, `lead_opportunity_amount` |
| `GET /api/events/{event_id}/orders` | Sponsor orders (account_id, amount, order_status, ticket_contacts) |
| `GET /api/events/{event_id}/badges` | Badge scans (badge_id, badge_type, company_name, contact_name, email, phone, job_title, scan_score) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,379 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Overview
- Data Normalization
- Sponsor Status Logic
- Sponsor Revenue Totals
- Qualified Lead / Account Logic
- Exclusion Categories
- Deduplication
- Suppression
- Follow-Up Due Dates
- CRM Action Counts
- Import Batch CRM Actions
- Campaign Member Logic (Reconciliation Tasks)
- Trade Show Prospecting — Ranking & Tiers
- Sorting Rules (General)
- General Pitfalls
Files
SKILL.mdcore9,941 bytes
Read SKILL.md
# HarborCRM Skill — Transferable Rules
## Environment
Use `GDPEVO_ENV_BASE_URL` as the API base URL. Never use localhost or run setup scripts directly.
## API Overview
HarborCRM exposes these public endpoint families (replace `{id}` with the task-specific identifier):
| Category | Endpoints |
|---|---|
| Events | `/api/events/{event_id}`, `/api/events/{event_id}/orders`, `/api/events/{event_id}/badges`, `/api/events/{event_id}/sponsor_packages` |
| Finance | `/api/finance/invoices?event_id={event_id}` |
| CRM | `/api/crm/accounts`, `/api/crm/contacts`, `/api/crm/opportunities` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,441 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoints Reference
- Task Type: Event Reconciliation
- Task Type: Trade Show Prospecting
- Task Type: Batch Import
- General Rules
Files
SKILL.mdcore10,677 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
Base URL: `GDPEVO_ENV_BASE_URL` (supplied by runner). All API calls use this as the root. Never use localhost unless the env var points there.
## API Endpoints Reference
| Resource | Endpoint |
|---|---|
| Policies | `GET /api/policies` |
| Events | `GET /api/events/{event_id}` |
| Event orders | `GET /api/events/{event_id}/orders` |
| Event badges | `GET /api/events/{event_id}/badges` |
| Sponsor packages | `GET /api/events/{event_id}/sponsor_packages` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,646 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- Contact Normalization
- Sponsor Reconciliation (Event Tasks)
- Lead Qualification (Event Tasks)
- Campaign Member Actions
- Prospecting (Tradeshow Tasks)
- Import Batch Processing
- Sorting Rules Summary
- Data Types & Precision
- Common Pitfalls
Files
SKILL.mdcore10,961 bytesSKILL.mdcore10,961 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
All API calls target the base URL supplied by the runner (e.g. `http://34.46.77.124:8001`). Do not use localhost or run setup scripts unless the remote URL itself points there.
## API Reference
| Purpose | Endpoint |
|---|---|
| Event details | `GET /api/events/{event_id}` |
| Sponsor orders | `GET /api/events/{event_id}/orders` |
| Sponsor packages | `GET /api/events/{event_id}/sponsor_packages` |
| Badge scans | `GET /api/events/{event_id}/badges` |
| Finance invoices | `GET /api/finance/invoices?event_id={event_id}` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,555 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Core API Endpoints
- Data Normalization
- Sponsor Status Determination
- Badge Classification & Lead Qualification
- Campaign Member Actions
- Badge-Only Contacts
- Prospecting / Exhibitor Qualification
- Batch Import Cleaning
- Exclusion Counts
- Sorting Rules (by output section)
- Key Pitfalls
Files
SKILL.mdcore11,185 bytesSKILL.mdcore11,185 bytesSKILL.mdcore11,185 bytesSKILL.mdcore11,185 bytes
Read SKILL.md
# HarborCRM Task Group Skill
## Environment
The shared HarborCRM API is at `${GDPEVO_ENV_BASE_URL}`. All data is read-only via GET endpoints. No write operations are needed or available.
## Core API Endpoints
Events, tradeshows, and import batches share a common pattern:
- `GET /api/events/{event_id}` — event metadata (dates, follow-up windows, lead opportunity amount, campaign code)
- `GET /api/events/{event_id}/orders` — sponsor orders (account, amount, order_status, ticket_contacts, voucher_code)
- `GET /api/events/{event_id}/badges` — badge scans (badge_type, company_name, contact_name, email, phone)
- `GET /api/events/{event_id}/sponsor_packages` — synonym for /orders
- `GET /api/finance/invoices?event_id={event_id}` — invoices (status, paid_amount, deferred_amount)
- `GET /api/tradeshows` — list of all trade showsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,623 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoints by Domain
- Sponsor Status Rules
- Badge Classification (Event Reconciliation Tasks)
- Campaign Member Actions (Event Reconciliation)
- Tradeshow Prospecting Qualification
- Import Batch Cleaning
- Follow-Up Dates
- Sorting Rules (apply everywhere)
- Data Type Conventions
- Key Pitfalls
Files
SKILL.mdcore11,028 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
Base URL: use the API base URL supplied by the runner. All endpoints are relative to this base. Do not start local services or run setup scripts.
## API Endpoints by Domain
### Events
- `GET /api/events/{event_id}` — event metadata (dates, follow-up offsets, lead opportunity amount, campaign code)
- `GET /api/events/{event_id}/orders` — sponsor orders (account_id, amount, order_status, package_level, ticket_contacts)
- `GET /api/events/{event_id}/badges` — scanned badges (badge_type, company_name, contact_name, email, phone, scan_score)
- `GET /api/events/{event_id}/sponsor_packages` — same shape as orders
### FinanceWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,515 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- Business Rules
- Contact Normalization
- CRM Matching
- Import Batch Processing
- Trade Show Prospecting
- Campaign Member Reconciliation
- Sorting Rules (Default)
- Output Conventions
- Common Pitfalls
Files
SKILL.mdcore14,951 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
Base URL: read from the runner or `environment_access.md`. The standard remote is `http://34.46.77.124:8001`. Never use localhost/127.0.0.1 unless the remote URL itself points there; `environment_access.md` overrides any localhost references in task prompts.
## API Reference
All endpoints are public GET. No auth required.
### Events
- `GET /api/events` — list all events
- `GET /api/events/{event_id}` — single event (includes `campaign_code`, `start_date`, `end_date`, `followup_days_after_end`, `sponsor_followup_days_after_end`, `lead_opportunity_amount`, `status`)
- `GET /api/events/{event_id}/orders` — sponsor orders for the event
- `GET /api/events/{event_id}/badges` — badge scans for the eventWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,649 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Endpoints by Task Type
- Data Models & Field Conventions
- Core Business Rules
- Sorting Rules Summary
- Common Pitfalls
Files
SKILL.mdcore15,131 bytes
Read SKILL.md
# HarborCRM Skill
## Environment
Base URL: `http://34.46.77.124:8001` (no auth required for public API). All `GET`.
## Endpoints by Task Type
### Event Reconciliation (train_001, train_004)
| Endpoint | Returns |
|---|---|
| `GET /api/events/{event_id}` | Event metadata, dates, campaign_code, followup_days, lead_opportunity_amount |
| `GET /api/events/{event_id}/orders` | Sponsor orders (same as sponsor_packages) |
| `GET /api/events/{event_id}/sponsor_packages` | Same as orders |
| `GET /api/events/{event_id}/badges` | Badge scans with contact info, badge_type, scan_score |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,301 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Normalization Conventions
- Sorting Rules (universal)
- API Endpoint Reference
- Task Pattern: Event Reconciliation
- Task Pattern: Trade-Show Prospecting
- Task Pattern: Import Batch Cleaning
- Common Pitfalls
Files
SKILL.mdcore22,090 bytes
Read SKILL.md
# HarborCRM Task Group Skill
## Environment
Use the base URL supplied by the runner or `environment_access.md`. Never use `localhost`, `127.0.0.1`, or run `env/setup.sh` — `environment_access.md` overrides all local references.
The API returns JSON. All outputs must be a single JSON object with no explanatory prose outside it.
## Normalization Conventions
- **Email**: lowercase, trim leading/trailing whitespace. Empty string when no email supplied.
- **Phone**: strip all non-digit characters (spaces, parens, dots, hyphens, leading `+1`). Empty string when no phone supplied.
- **USD amounts**: always integers (whole dollars, no cents).
- **Dates**: `YYYY-MM-DD` format. Compute follow-up dates as `end_date + N days` (calendar addition).
- **Booleans**: JSON `true`/`false`, not strings.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,163 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment entrypoint
- Task taxonomy
- Cross-cutting conventions
- Common pitfalls
- API call order (all patterns)
Files
SKILL.mdcore12,511 bytesSKILL.mdcore12,511 bytes
Read SKILL.md
# MedBridge Sales Ops — Quote, Reconciliation & Account Decision Packages
## Environment entrypoint
The MedBridge Sales Ops API is always accessed through the task environment base URL.
When the task text mentions `API_BASE_URL`, `BASE_URL`, `localhost`, `127.0.0.1`, or
`env/setup.sh`, substitute the remote URL declared in the runner's environment access
configuration. Do **not** start a local env or read an `env/` source directory.
---
## Task taxonomy
MedBridge tasks fall into three patterns. Identify which one applies before calling
any API endpoint.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,346 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API surface
- Core procedures
- Output format & conventions
- Field mapping reference
- Common pitfalls
- Entity relationship map
Files
SKILL.mdcore20,015 bytes
Read SKILL.md
# MedBridge Sales Ops — Quote & Engagement Reconciliation
## Environment
The MedBridge Sales Ops API is a read-only JSON API that provides CRM, quote,
logistics, product catalog, and milestone engagement data. The runner supplies
the base URL as `API_BASE_URL` or `BASE_URL`. Construct all requests as
`GET {base_url}/api/...`.
## API surface
| Method | Path | Purpose |
|--------|------|---------|
| `GET` | `/api` | List all collections |
| `GET` | `/api/customers` | All customer records |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,252 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Task Type 1: Quote Revision with Freight
- Task Type 2: Indicative RFQ / Module Quote (No Freight)
- Task Type 3: Opportunity Reconciliation with Event/Voucher
- Task Type 4: Engagement Reconciliation (Full Accounting View)
- Cross-Cutting Business Rules
- Common Pitfalls
- Response Format
Files
SKILL.mdcore20,354 bytesSKILL.mdcore20,354 bytes
Read SKILL.md
# MedBridge Sales Ops — Condition Fewshot Skill
## Environment
```
API_BASE_URL=http://34.46.77.124:8002
```
All task prompts reference `API_BASE_URL`, `BASE_URL`, or a localhost placeholder. Always resolve to the remote base above. The API is a read-only shared data source with 12 collections.
### API Endpoints
| Method | Path | Purpose |
|--------|------|---------|
| GET | `/api` | List all collections and endpoints |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,199 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Overview
- Data Source Precedence
- Product Pricing Rules
- Freight & Transport Rules
- Payment Terms Rules
- Reconciliation Rules (Opportunities, Invoices, Revenue)
- Event & Voucher Rules
- Common Pitfalls
- Output Field Conventions
Files
SKILL.mdcore18,424 bytesSKILL.mdcore18,424 bytes
Read SKILL.md
# MedBridge Sales Ops API — Skill for Quote & Reconciliation Packages
## Environment
All API calls go to the shared MedBridge Sales Ops base URL. The task runner provides it
as `API_BASE_URL`, `BASE_URL`, or the task environment base URL. There is no local server;
the remote API is the only data source.
## API Overview
| Collection | Key endpoints |
| ---------------- | -------------------------------------------------------------------- |
| customers | `GET /api/customers`, `GET /api/customers/<id>` |
| products | `GET /api/products`, `GET /api/products/<code>` |
| quotes | `GET /api/quotes`, `GET /api/quotes/<id>` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,609 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API conventions
- Quote workflows
- Module RFQ workflow (EXW only, no freight)
- Opportunity & engagement reconciliation
- Source precedence
- Common pitfalls
Files
SKILL.mdcore13,821 bytesSKILL.mdcore13,821 bytes
Read SKILL.md
# MedBridge Sales Ops — Quote & Engagement Reconciliation
## Overview
The MedBridge Sales Ops API serves shared CRM, product-catalog, quoting, freight,
invoice, payment, revenue-recognition, event, and voucher data. Use it to build
quote decision packages and account-ready reconciliations from a single source of
truth. Every response field that names an ID, code, date, amount, or controlled
status must come from API records — never invent identifiers or infer statuses
from task narrative alone.
---
## API conventions
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,763 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API access pattern
- Catalog pricing: tier selection
- Freight: finding the right records
- Freight risk and recommendations
- Customer payment terms
- Quote scope rules
- Milestone reconciliation pattern
- Accounting and collection actions
- Event and voucher handling
- Common pitfalls
- End-to-end workflow
- Output conventions
Files
SKILL.mdcore14,591 bytesSKILL.mdcore14,591 bytes
Read SKILL.md
# MedBridge Sales Ops API — Quote, Reconciliation & Engagement Skill
## API access pattern
All data lives at `{BASE_URL}/api/{collection}`. The API returns full collections
with a `records` array — dump the whole list and filter client-side. There are
no query parameters or server-side filters.
### Available collections
| Collection | What it holds |
|---|---|
| `customers` | Customer records with contacts, payment profiles, regions, segments |
| `quotes` | Quote headers, line items, quantities, incoterms, revision notes |
| `rfqs` | Inbound RFQs with requested modules, quantities, narrative instructions |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,973 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Workflow 1: Quote + Freight Decision Package (Tasks 001, 004)
- Workflow 2: RFQ / Indicative Module Quote (Task 002)
- Workflow 3: Account Reconciliation + Milestone Revenue (Tasks 003, 005)
- Cross-cutting Rules
- Quick-start recipe per task pattern
Files
SKILL.mdcore22,748 bytesSKILL.mdcore22,748 bytes
Read SKILL.md
# MedBridge Sales Ops API — Transferable SOP
## Environment
Base URL comes from the runner via `API_BASE_URL`, `BASE_URL`, or the task environment.
The API is read-only (GET only). All responses are JSON. No authentication is required
in the evaluation harness. The API's `/health` endpoint confirms the service name
`"MedBridge Sales Ops"` and returns the seed.
### Endpoint catalog
```
GET /api — list all collections and endpoints
GET /api/customers — all customers
GET /api/customers/<customer_id>When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,256 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment & API Entrypoint
- 2. API Collections & Navigation Pattern
- 3. Product Catalog & Price Tier Resolution
- 4. Freight Quote Filtering & Validation
- 5. Route Risk & Mode Recommendation
- 6. Customer Record & Payment Terms
- 7. Quote Types & Scope Rules
- 8. Opportunity Reconciliation (Tasks 003, 005)
- 9. Common Pitfalls & Edge Cases
- 10. Task-to-API Mapping Summary
- 11. Output Structure Rules
- 12. API Behavior Notes
Files
SKILL.mdcore16,641 bytes
Read SKILL.md
# MedBridge Sales Ops API — Transferable SKILL.md
## 1. Environment & API Entrypoint
- The API base URL is provided by the runner (env, `API_BASE_URL`, `BASE_URL`, `http://34.46.77.124:8002`, etc.). Always use the runner-provided URL.
- All endpoints are read-only `GET`. No authentication required in this staging environment.
- **Root discovery**: `GET /api` returns the collection list and all available endpoints.
## 2. API Collections & Navigation Pattern
| Collection | Endpoint | Join Keys |
|---|---|---|
| `customers` | `/api/customers`, `/api/customers/<id>` | `id` |
| `products` | `/api/products`, `/api/products/<code>` | `code` |
| `rfqs` | `/api/rfqs`, `/api/rfqs/<id>` | `id`, `customer_id` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,776 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoint Inventory
- Task Dispatch: Quote vs. RFQ vs. Reconciliation
- Catalog Tier Selection (Quantity Brackets)
- EXW Total Calculation
- Freight Record Filtering (Critical — Many Distractors)
- Freight Validity and Reconfirmation
- Risk Assessment
- Recommended Transport Mode
- Payment Terms Assignment
- Quote Scope (EXW vs. Freight-Included)
- Offer Validity
- Customer Name Resolution
- Opportunity Reconciliation Rules
- Known Distractor Patterns
- Date Conventions
- Output Format
- Quick Reference: Field Sources
Files
SKILL.mdcore14,803 bytes
Read SKILL.md
# MedBridge Sales Ops API — Transferable SOP
## Environment
All API calls go to `{API_BASE_URL}/api/` where `API_BASE_URL` is provided by the
task runner. When task text mentions `localhost`, `127.0.0.1`, `BASE_URL`, or
`env/setup.sh`, substitute the remote base URL from the runner instead.
## API Endpoint Inventory
| Method | Path | Returns |
|--------|------|---------|
| GET | `/api` | Collection index + endpoint listing |
| GET | `/api/customers` | All customers |
| GET | `/api/customers/<id>` | Single customer |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,255 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- Decision SOPs
- Output Conventions
- Common Pitfalls
Files
SKILL.mdcore12,554 bytes
Read SKILL.md
# CRM Support Console — Agent Skill
## Environment
- **Base URL**: Read `environment_access.md` first. If it specifies `GDPEVO_ENV_BASE_URL`, use that. Otherwise default to the URL in the task prompt.
- **Health check**: `GET /health` → `{ok, service}`
- **Catalog**: `GET /api/catalog` lists all endpoints and record counts (generated seed data).
---
## API Reference
### Service Tickets (internet / voice / video)
| Endpoint | Key Fields |
|---|---|
| `GET /api/tickets/<id>` | `ticket_id`, `account_id`, `service_area`, `service_type`, `subscribed_mbps`, `status`, `issue_summary` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,364 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Base URL
- API Reference
- Resolution SOP — Ticket Triage (train_001 / train_004)
- Resolution SOP — Mobile Case Routing (train_002)
- Resolution SOP — Mobile Data Recovery (train_005)
- Enterprise Incident Response SOP (train_003)
- Output Conventions (ALL task types)
- Common Pitfalls
- Quick Pre-Flight Checklist
Files
SKILL.mdcore13,344 bytes
Read SKILL.md
# CRM Support Console — Solver Skill
## Base URL
Always use the remote base URL from `environment_access.md`. **Never** use `localhost`, `127.0.0.1`, or any URL embedded in a task prompt — the `environment_access.md` override is authoritative.
## API Reference
| Endpoint | Use |
|---|---|
| `GET /health` | Confirm API is reachable before starting |
| `GET /api/catalog` | Discover available endpoints / schemas |
| `GET /api/tickets/{ticket_id}` | Ticket status, account link, reported issue, service type |
| `GET /api/accounts/{account_id}` | Account standing: active, suspended, on-hold, credit class |
| `GET /api/diagnostics/{ticket_id}` | Latency, stability, bandwidth metrics; returns boolean flags |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,767 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Base URL
- Core API Reference
- Ticket Triage Rules (Internet / Video / Voice)
- Mobile Case Queue Rules
- Enterprise Incident Response Rules
- SLA Queue Review Rules (Train 004 variant)
- Output Conventions (Checklist)
- Common Pitfalls
- Compact SOP
Files
SKILL.mdcore11,103 bytes
Read SKILL.md
# CRM Support Console — Reusable Skill
## Base URL
Use the URL from `environment_access.md`. Ignore `127.0.0.1` / `localhost` in task prompts.
Known public base: `http://34.46.77.124:8003` (verify against environment_access.md).
---
## Core API Reference
| Endpoint | Use |
|---|---|
| `GET /api/tickets/<id>` | Ticket detail: status, account_id, service_type, area |
| `GET /api/accounts/<id>` | Account status (active / suspended / hold), service_area |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,421 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoint Reference
- Business Rules by Domain
- Common Pitfalls
- Compact SOP
Files
SKILL.mdcore10,243 bytes
Read SKILL.md
# Support Console API Skill
## Environment
- Base URL provided by `environment_access.md` (GDPEVO_ENV_BASE_URL). Never use localhost unless the remote URL explicitly points there.
- Always consult `/api/catalog` first when entering an unfamiliar task category to confirm available endpoints.
## API Endpoint Reference
### Core Endpoints
| Endpoint | Purpose |
|---|---|
| `/api/accounts/<id>` | Account status, auth (last_login_status, account_recovery_status), tier, service_area |
| `/api/tickets/<id>` | Ticket detail: account_id, service_area, service_type, subscribed_mbps, status |
| `/api/outages?service_area=<area>` | Active outages with outage_id, eta_hours, impact_score, service_types affected |
| `/api/diagnostics/<ticket_id>` | Post-hoc root causes, bandwidth/latency/jitter measurements |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,069 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Habits
- Business Rule Reference
- Output Field Conventions
- Compact SOP
- Common Pitfalls
Files
SKILL.mdcore8,459 bytes
Read SKILL.md
# Support Console Solver — SKILL.md
## Overview
This skill covers the task_group_003 support-console API used to resolve service tickets, mobile cases, enterprise incidents, and queue-quality reviews. The API is read-only and returns all evidence needed to construct structured JSON answer payloads.
## API Usage Habits
### Base URL & Discovery
- Use the remote URL provided by the harness (override any localhost references in task prompts).
- Start with `/health` to confirm connectivity, then `/api/catalog` for the full endpoint list and record counts.
- Endpoint list: `/api/accounts[/<id>]`, `/api/tickets[/<id>]`, `/api/outages?service_area=<area>`, `/api/diagnostics/<ticket_id>`, `/api/troubleshooting/<ticket_id>`, `/api/cases[/<id>]`, `/api/customers`, `/api/lines[/<id>]`, `/api/devices[/<id>]`, `/api/plans[/<id>]`, `/api/bills`, `/api/enterprise/incidents[/<id>]`, `/api/enterprise/export-runs?incident_id=<id>`, `/api/enterprise/messages?query=<text>`, `/api/enterprise/sla/<enterprise_account_id>`.
### Data Fetching Pattern
- For batch tasks, fetch tickets/cases first to get IDs (account_id, line_id, device_id, service_area), then fan out to accounts, diagnostics, outages, lines, devices in parallel.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,923 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Base
- Core SOP: Always Fetch the Full Record Chain
- Wireline Ticket Resolution Rules
- Mobile Case Resolution Rules
- Enterprise Incident Response Rules
- Output Format Conventions
- Common Pitfalls
Files
SKILL.mdcore7,673 bytes
Read SKILL.md
# Support Console Skill — SOP for Ticket/Case Resolution
## API Base
Use the harness-provided base URL (not localhost). All endpoints listed at `/api/catalog`.
## Core SOP: Always Fetch the Full Record Chain
For any ticket or case ID found in a payload, retrieve every linked record before deciding:
- **Wireline tickets**: `/api/tickets/<id>` → `/api/accounts/<account_id>` → `/api/diagnostics/<id>` → `/api/troubleshooting/<id>` → `/api/outages?service_area=<area>`
- **Mobile cases**: `/api/cases/<id>` → `/api/lines/<line_id>` → `/api/devices/<device_id>` → `/api/plans/<plan_id>` → `/api/bills` (filter by customer_id)
- **Enterprise incidents**: `/api/enterprise/incidents/<id>` → `/api/enterprise/accounts` (filter by name) → `/api/enterprise/export-runs?incident_id=<id>` → `/api/enterprise/messages?query=<keyword>` → `/api/enterprise/sla/<enterprise_account_id>`
## Wireline Ticket Resolution Rules
### Outage-First RuleWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,730 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- ID Prefix Glossary
- Endpoint Catalog
- Data Models
- Standard Operating Procedures
- Common Pitfalls
- Output Conventions
- Data Generation Notes
Files
SKILL.mdcore13,948 bytes
Read SKILL.md
# Support Console Skill — Task Group 003
## Environment
- **Base URL**: `http://34.46.77.124:8003` (remote only; never use localhost)
- **Service**: `task_group_003_support_console`
- **Auth**: None required (public API)
- **Data shape**: All timestamps are ISO-8601 with `Z` suffix. All enums are `UPPER_SNAKE_CASE`. All IDs are prefixed (see ID glossary below).
## ID Prefix Glossary
| Prefix | Entity | Example |
|--------|--------|---------|
| `ACC-` | Consumer/Business Account | `ACC-5107` |
| `TCK-` | Support Ticket | `TCK-5107` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,744 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Quick SOP: Diagnose Any Issue
- API Reference
- Error Handling Conventions
- Entity Relationships
- ISP Ticket Root Cause → Resolution Map
- Mobile Case Diagnostic Decision Tree
- Enterprise Incident SOP
- Common Pitfalls
- Effective Query Patterns
- Output Conventions
- Distinguishing Canonical from Generated Data
Files
SKILL.mdcore14,402 bytesSKILL.mdcore14,402 bytes
Read SKILL.md
# Support Console API — Solver Skill
## Overview
This is a **telecom/ISP support console** with three domains connected by shared identifiers:
- **Residential/Business ISP** — tickets, accounts, outages, diagnostics, troubleshooting
- **Mobile** — cases, customers, lines, devices, plans, bills
- **Enterprise** — enterprise accounts, incidents, export runs, messages, SLA contracts
Base URL: read `environment_access.md` — it overrides any localhost references in task text.
## Quick SOP: Diagnose Any Issue
1. Read the task prompt. Identify which domain(s) are involved.
2. Pull the primary record (ticket, case, incident) from its list or detail endpoint.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,807 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Base URL
- API Catalog & Error Convention
- Data Model Relationships
- Task Type 1: Ticket Batch Resolution
- Task Type 2: Mobile Contact-Center Queue
- Task Type 3: Enterprise Export Incident Response
- Task Type 4: Queue Quality Analysis (Pre-SLA Review)
- Task Type 5: Mobile Data Recovery
- Common Pitfalls
- Field Value Quick Reference
Files
SKILL.mdcore14,811 bytes
Read SKILL.md
# Support Console — Operations Analyst SKILL
## Base URL
Always read `environment_access.md` first. It overrides any `localhost` or `127.0.0.1` text in task prompts. The canonical base is the `GDPEVO_ENV_BASE_URL` from that file.
## API Catalog & Error Convention
**Catalog**: `GET /api/catalog` lists all endpoints and record counts. Use for discovery.
**Error convention**: Missing/invalid records return `{"error": "not_found"}` with HTTP 200 — never a 404. Always check for the `"error"` key, not the HTTP status code.
**Key endpoints**:
| Endpoint | Description |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,065 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Fundamentals
- Data Cleaning Rules
- Output Precision Rules
- Controlled Vocabularies (Enums)
- Business Rules
- Sorting Conventions
- JSON Output Rules
- Workflow Pattern
- Common Pitfalls
Files
SKILL.mdcore15,933 bytesSKILL.mdcore15,933 bytesSKILL.mdcore15,933 bytesSKILL.mdcore15,933 bytes
Read SKILL.md
# ApexCloud Retention Operations — Reusable Skill
## Overview
This skill covers the ApexCloud Retention Operations API for customer success workflows: renewal risk scoring, QBR metrics compilation, receivables/pipeline reviews, churn model validation, and retention action boards. Every task follows a **fetch → compute → fill-template → return-JSON** pattern. Output is always a single JSON object with no markdown wrapping.
---
## API Fundamentals
### Base URL
Use the remote URL from `environment_access.md`. Never use `localhost`, `127.0.0.1`, or URLs from task prompts unless they match the remote URL.
### Endpoint ReferenceWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,294 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoint Reference
- Numeric Precision Rules
- Controlled Vocabularies (Enums)
- Business Logic Rules
- Workflow Procedure
- Pitfalls & Edge Cases
Files
SKILL.mdcore12,919 bytes
Read SKILL.md
# ApexCloud Retention Operations API — Reusable Skill
## Environment
- **Base URL**: `http://34.46.77.124:8004`
- All task prompts may reference `localhost:8074` or `127.0.0.1:8074`; ignore those — always use the base URL above.
- Never run `env/setup.sh`, never use localhost/127.0.0.1.
---
## API Endpoint Reference
These endpoint families are available. Use the one that matches the task's data requirements.
### Account & MetricsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,643 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment
- 2. API Endpoint Catalogue
- 3. Output Precision Conventions
- 4. Controlled Vocabulary
- 5. Business-Policy Rules
- 6. Execution Workflow
- 7. Common Pitfalls
Files
SKILL.mdcore13,972 bytesSKILL.mdcore13,972 bytesSKILL.mdcore13,972 bytes
Read SKILL.md
# ApexCloud Retention Operations — Agent Skill
Use this skill to execute revenue-retention analysis tasks against the ApexCloud Retention Operations API. It covers the API surface, output conventions, controlled vocabularies, business-policy rules, and common pitfalls distilled from solved Q2/Q3 2026 tasks.
---
## 1. Environment
- **Base URL:** Consult `environment_access.md` in the task directory. Use the `GDPEVO_ENV_BASE_URL` value (a remote IP) for all API calls.
- **Never** start the task-group environment, run `env/setup.sh`, or use `localhost`/`127.0.0.1` directly — the environment_access.md URL is authoritative and overrides any localhost reference in a task prompt.
---
## 2. API Endpoint Catalogue
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,984 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Conventions
- Data Extraction Rules
- Output Field Conventions
- Business-Policy Rules
- Exclusion Rules & Pitfalls
Files
SKILL.mdcore12,202 bytesSKILL.mdcore12,202 bytesSKILL.mdcore12,202 bytes
Read SKILL.md
# ApexCloud Retention Operations — Skill Guide
## Overview
This skill covers the ApexCloud Retention Operations API: account profiling, metrics, support-ticket hygiene, NPS, A/R aging, billing, CRM pipeline, HR context, event operations, and churn-model exports. Use it to build renewal risk queues, QBR packets, receivables reviews, churn-model readouts, and retention action boards.
---
## API Conventions
### Base URL
Use the remote environment base URL. Do **not** use `localhost`, `127.0.0.1`, or any `env/setup.sh` reference.
### Endpoint FamiliesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,905 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Base URL
- Core Endpoint Catalogue
- Controlled Vocabularies
- Precision Rules
- Business-Policy Rules
- Exclusion Rules
- Common Pitfalls
- Workflow Pattern
Files
SKILL.mdcore15,773 bytesSKILL.mdcore15,773 bytesSKILL.mdcore15,773 bytesSKILL.mdcore15,773 bytesSKILL.mdcore15,773 bytes
Read SKILL.md
# ApexCloud Retention Operations — Skill Reference
## Overview
This skill covers the ApexCloud Retention Operations API, a multi-source customer-success data platform. Tasks span renewal-risk queueing, QBR metrics packets, receivables-and-pipeline operations reviews, churn-model validation with outreach ranking, and high-touch retention action boards.
---
## API Base URL
Always read `environment_access.md` first. Use the `GDPEVO_ENV_BASE_URL` from that file as the API root. Never use `localhost`, `127.0.0.1`, or paths from `env/setup.sh` unless the environment-access file itself points there. Do not start the service locally.
---
## Core Endpoint CatalogueWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,557 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API Endpoint Reference
- 2. ARR / Revenue Source Convention
- 3. CRM / AR Customer Matching
- 4. NPS Conventions
- 5. Clean Ticket Counting
- 6. Overdue Balance Calculation
- 7. Pipeline Analysis
- 8. Risk Assessment Framework
- 9. Churn Model Methodology
- 10. Output Precision Rules (Deterministic)
- 11. Follow-Up Calendar Convention
- 12. Portfolio / Segment Summary Construction
- 13. QBR Metric Packet Construction
- 14. Common Pitfalls
- 15. Controlled Vocabulary Quick Reference
Files
SKILL.mdcore14,241 bytesSKILL.mdcore14,241 bytesSKILL.mdcore14,241 bytesSKILL.mdcore14,241 bytes
Read SKILL.md
# ApexCloud Retention Operations — Reusable Skill
## 1. API Endpoint Reference
All data comes from the ApexCloud Retention Operations API. Key endpoints:
| Endpoint | Query Params | Returns |
|---|---|---|
| `/api/accounts` | none | All account profiles (billing_arr_current, crm_arr, tenure, renewal_date, lifecycle_status, segment, region, aliases) |
| `/api/accounts/{id}` | none | Single account profile |
| `/api/accounts/{id}/metrics` | `?start=YYYY-MM&end=YYYY-MM` | Monthly recognized_revenue, nps_score, product_usage, sla_compliance, support_ticket_count, active_seats, survey_status |
| `/api/accounts/{id}/tickets` | `?start=YYYY-MM-DD&end=YYYY-MM-DD` | Support tickets with SLA fields, spam/duplicate flags, severity |
| `/api/accounts/{id}/nps` | `?start=YYYY-MM-DD&end=YYYY-MM-DD` | Individual NPS survey responses with retracted flag and survey_channel |
| `/api/finance/ar-aging` | none | AR aging by quarter for all customers (1_30, 31_60, 61_90, 90_plus, current buckets) |
| `/api/billing/snapshots` | `?start=YYYY-MM-DD&end=YYYY-MM-DD` | Quarterly billing ARR snapshots per account with source field |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,307 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Conventions
- Data Models & Field Reference
- Business Rules
- Output Conventions
- Churn Model SOP
- Common Workflow Patterns
- Pitfalls
Files
SKILL.mdcore17,544 bytesSKILL.mdcore17,544 bytes
Read SKILL.md
# ApexCloud Retention Operations — Self-Serve Analytics
## Overview
This skill covers self-serve analytics against the **ApexCloud Retention Operations API** for customer success, revenue operations, and churn analytics workflows. It applies to risk queue builds, QBR metric packets, receivables reviews, churn model validation, and retention action boards.
---
## API Conventions
### Base URL
Always use the remote base URL supplied via `environment_access.md` in the workspace. Never use `localhost`, `127.0.0.1`, or URLs from task prompts — `environment_access.md` overrides them.
### Endpoint inventoryWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,050 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Endpoint Reference
- Data Retrieval Patterns
- Controlled Vocabularies
- Business Rules
- Output Format Conventions
- Common Pitfalls
- Workflow Checklist
Files
SKILL.mdcore16,502 bytesSKILL.mdcore16,502 bytesSKILL.mdcore16,502 bytes
Read SKILL.md
# ApexCloud Retention Operations — Self-Service Analytics Skill
## Overview
This skill covers the ApexCloud Retention Operations API and the standard workflows for building retention risk queues, QBR metric packets, receivables reviews, churn model validations, and high-touch retention action boards. All tasks share the same API surface, controlled vocabularies, and output conventions described below.
**Base URL:** Use the remote API URL provided in `environment_access.md`. Never use localhost unless the remote URL itself redirects there.
---
## API Endpoint Reference
### Accounts
| Endpoint | Method | Key Parameters | Returns |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,585 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- How to use this skill
- 1. Base URL
- 2. API Endpoint Catalog
- 3. Data-Cleaning Rules (apply in every task)
- 4. Common Pitfalls
- 5. Business-Logic Conventions
- 6. Output Precision and Formatting
- 7. Controlled Vocabularies
- 8. Policy Codes
- 9. Task-Type Quick Reference
- 10. Execution Checklist
Files
SKILL.mdcore21,387 bytesSKILL.mdcore21,387 bytesSKILL.mdcore21,387 bytesSKILL.mdcore21,387 bytesSKILL.mdcore21,387 bytesSKILL.mdcore21,387 bytes
Read SKILL.md
# ApexCloud Retention Operations — Reusable SOP
## How to use this skill
This skill covers the ApexCloud Retention Operations API (`GDPEVO_ENV_BASE_URL`). It documents endpoints, data-cleaning rules, business-logic conventions, output vocabularies, and precision requirements that apply across all retention workflows (risk queues, QBR packets, receivables reviews, churn validation, and high-touch boards). Read it once, then follow the rules for every task variant.
---
## 1. Base URL
Use the URL from `environment_access.md`:
```
GDPEVO_ENV_BASE_URL=http://34.46.77.124:8004
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,820 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Available API Endpoints
- General Conventions
- Domain 1: Claims & AP Reimbursement Close (Train 001, 004)
- Domain 2: Vendor Onboarding & Compliance Release (Train 002, 005)
- Domain 3: Prepaid Expense Close (Train 003)
- Cross-Cutting Workflow
Files
SKILL.mdcore16,181 bytesSKILL.mdcore16,181 bytesSKILL.mdcore16,181 bytesSKILL.mdcore16,181 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill
## Environment
All API calls use the shared remote base URL from `environment_access.md`:
`http://34.46.77.124:8005`. Never use localhost, 127.0.0.1, or local env scripts.
## Available API Endpoints
Every endpoint is available both with and without the `/api/` prefix:
| Short Path | Long Path | Returns |
|---|---|---|
| `/claims` | `/api/claims` | Expense claim records (id, status, amount, vendor, submitter, etc.) |
| `/bills` | `/api/ap/bills` | AP bill records (id, claim_id, status, amount, vendor_id, etc.) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,028 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence (Critical)
- General Conventions
- Workflow 1: Reimbursement Claim Close Review
- Workflow 2: Vendor Onboarding Finance-Risk Review
- Workflow 3: Prepaid Expense Close Reconciliation
- Workflow 4: Stale AP Snapshot Reconciliation
- Workflow 5: Post-Account-Change Payment Release
- Cross-Cutting Rules
Files
SKILL.mdcore20,849 bytesSKILL.mdcore20,849 bytesSKILL.mdcore20,849 bytes
Read SKILL.md
# ERP Finance Expense-Control — Reusable Skill
## Environment
All tasks use a shared remote ERP finance API. The base URL is provided by the
runner (see `environment_access.md` at the task root). Do not use localhost,
127.0.0.1, or local setup scripts. Always query the remote API for current
state.
Public endpoints (prefer the `/api/…` variant when both exist):
| Domain | Endpoints |
|---|---|
| Claims | `/claims`, `/api/claims` |
| AP Bills | `/bills`, `/api/ap/bills` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,030 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & API
- Workflow 1: Expense Claim Reimbursement-to-AP Close Review
- Workflow 2: Vendor Onboarding Finance-Risk Release
- Workflow 3: Prepaid Expense Close Check
- Workflow 4: Stale AP Snapshot Reconciliation
- Workflow 5: Account-Change Payment Release Review
- General Conventions
Files
SKILL.mdcore16,032 bytes
Read SKILL.md
# ERP Finance Expense-Control — Reusable SKILL.md
## Environment & API
- **Base URL**: `http://34.46.77.124:8005` (use `/api/*` variants).
- **Filtering**: Exact-match query parameters by field name (e.g. `?claim_id=CLM-2025-0001`, `?business_id=BUS-2025-0009`, `?account=1250&period=2025-03`).
- **Pagination**: `limit` and `offset` query params.
- **Key endpoints**:
- `/api/claims` — expense claims (status, amount, vendor_id, policy_flags, receipt_status)
- `/api/ap/bills` — AP bills (status, amount, claim_id, vendor_id, bill_id, account)
- `/api/ap/payments` — payments (status, amount, bill_id, payment_id, method)
- `/api/vendors` — vendor master data (status, bank_account_last4, tax_id, legal_name)
- `/api/compliance/objects` — compliance records (bank_account_status, pep_status, license_expiry, risk_score, sanctions_check_status, shell_company_suspected, missing_fields, ubo_list, tax_id)
- `/api/prepaids/invoices` — prepaid schedules (monthly_amortization, original_amount, service_start/end, data_quality_flags, account)
- `/api/prepaids/gl-balances` — GL ending balances by account+period+entityWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,351 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & API
- 1. Expense Claim Close Review (Reimbursement-to-AP)
- 2. Vendor Onboarding Finance-Risk Release
- 3. Prepaid Close Check
- 4. Stale AP Snapshot Reconciliation
- 5. AP Payment Release After Account-Change Events
- General Conventions
Files
SKILL.mdcore14,139 bytes
Read SKILL.md
# ERP Finance Expense-Control — Skill Reference
## Environment & API
All tasks use the shared ERP finance REST API. Filter records with exact-match query
parameters by field name (`claim_id`, `business_id`, `vendor_id`, `account`, `period`,
`bill_id`). Paginate with `limit` and `offset`. The API is read-only; no
create/update/delete operations are needed.
Key endpoints:
| Endpoint | Use |
|---|---|
| `/api/claims` | Expense claim records |
| `/api/ap/bills` | AP bills linked to claims |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,729 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Core Endpoints
- Expense Claim Close Review Workflow
- Vendor Onboarding Finance-Risk Review
- Prepaid-Expense Close Reconciliation
- Stale AP Snapshot Reconciliation
- Account-Change Payment Release Review
- General Conventions (All Tasks)
Files
SKILL.mdcore15,794 bytesSKILL.mdcore15,794 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill
## Overview
This skill covers expense claim close review, vendor onboarding risk assessment,
prepaid-expense close reconciliation, stale AP snapshot correction, and
account-change payment-release review in the shared ERP finance API.
**API base URL:** Use the runner-provided base URL. All endpoints support both
`/api/<resource>` and `/<resource>` paths with exact-match query parameters,
`limit`, and `offset` for pagination.
## Core Endpoints
| Resource | Endpoints | Key Fields |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,230 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Overview
- 1. Expense Claim Close Review (AP Reimbursement Batch)
- 2. Vendor Onboarding Finance-Risk Review
- 3. Prepaid Expense Close Reconciliation
- 4. Stale AP Snapshot Reconciliation
- 5. AP Payment Release After Account-Change Events
- General Conventions
Files
SKILL.mdcore17,521 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill
## API Overview
Base the API entrypoint on the URL provided by the task runner. The ERP finance environment exposes these endpoints used across expense-control tasks:
| Endpoint | Purpose |
|---|---|
| `/api/claims` | Expense claim records |
| `/api/ap/bills` | AP bill/invoice records |
| `/api/ap/payments` | Payment records linked to bills |
| `/api/vendors` | Vendor master data (bank, tax ID, status) |
| `/api/compliance/objects` | Compliance & risk screening per business |
| `/api/prepaids/invoices` | Prepaid invoice schedules |
| `/api/prepaids/gl-balances` or `/gl/balances` | GL ending balances by account/period |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,335 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Reference
- Workflow 1: Claims-to-AP Close Review
- Workflow 2: Vendor Onboarding Release Control
- Workflow 3: Prepaid Close Reconciliation
- Workflow 4: Stale AP Snapshot Reconciliation
- Workflow 5: Account-Change Payment Release
- Cross-Cutting Conventions
Files
SKILL.mdcore21,418 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill
## Overview
This skill covers five expense-control workflows in the ERP finance domain:
1. **Claims-to-AP close review** — classify reimbursement claims against AP bills/payments
2. **Vendor onboarding release control** — compliance screening for vendor access decisions
3. **Prepaid close reconciliation** — schedule-to-GL variance analysis with straight-line amortization
4. **Stale AP snapshot reconciliation** — reconcile candidate claims against current ERP state vs stale export
5. **Account-change payment release** — compliance-gated payment release after vendor account-change events
All workflows use a shared ERP finance API. **Always query the API for current data** — never rely on stale snapshots or local files as the system of record.
---
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,218 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Reference
- Workflow 1: Reimbursement-to-AP Close Review
- Workflow 2: Vendor Onboarding Finance-Risk Release Control
- Workflow 3: Prepaid Expense Close Check
- Workflow 4: Stale AP Export Reconciliation
- Workflow 5: AP Payment Release After Account-Change Events
- Output Conventions (All Workflows)
- Common Pitfalls
- Workflow Quick Reference
Files
SKILL.mdcore28,275 bytesSKILL.mdcore28,275 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill
## Overview
This skill covers five expense-control workflows in the task_group_005 ERP finance
environment. All data is accessed through a shared JSON API. The skill describes
transferable business rules, API usage patterns, output conventions, calculation
rules, source precedence, and common pitfalls for held-out tests.
## API Reference
### Base URL
Use the environment-provided base URL. The remote environment serves both the
prefixed (`/api/...`) and unprefixed (`/...`) paths for every endpoint.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,421 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment
- Workflow 1: Claims Close Review (Reimbursement-to-AP)
- Workflow 2: Vendor Onboarding Risk Review
- Workflow 3: Prepaid Close Reconciliation
- Workflow 4: Stale AP Snapshot Reconciliation
- Workflow 5: Account-Change Payment Release Review
- Cross-Cutting Conventions
- End-to-End Workflow Checklist
Files
SKILL.mdcore22,625 bytes
Read SKILL.md
# ERP Finance Expense-Control — Reusable Skill
## Overview
This skill covers five related expense-control workflows in a shared ERP finance environment:
1. **Claims close review** — classify reimbursement claims as paid, payable, or blocked
2. **Vendor onboarding risk review** — make release-control decisions with compliance checks
3. **Prepaid expense close reconciliation** — reconcile prepaid schedules against GL balances
4. **Stale AP snapshot reconciliation** — correct stale AP exports against current ERP data
5. **Account-change payment release review** — assess risk for payment releases after vendor account changes
All workflows share the same API base, common sorting/output conventions, and a consistent source-precedence rule.
---
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,663 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment
- 2. Source Precedence
- 3. ID Conventions
- 4. Output Format Rules (Apply to Every Task)
- 5. Reusable Business Rules
- 6. Arithmetic Checks (Post-Computation Validation)
- 7. Output Schema Pitfalls
- 8. Task-Type Quick Index
- 9. Exclusion Rules
Files
SKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytes
Read SKILL.md
# ProcureOps Task Group 006 — Solver SKILL
## 1. Environment
- **Base URL**: `http://34.46.77.124:8006` (from `environment_access.md`). Never use localhost/127.0.0.1.
- **Endpoints**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- Accept JSON responses; all IDs and status fields come from the API, not from task payloads.
## 2. Source Precedence
1. **ProcureOps API** — authoritative for all record values (statuses, amounts, dates, quantities, supplier names, risk ratings, approval states).
2. **Task payload** (memo/JSON packet) — provides target IDs (which records to review), business-control parameters (tax rate, currency, opening balance assumptions), and any local registers (chargeback register). Use these to scope your API queries, *not* to override API values.
3. **Answer template** — defines the output shape, enum sets, sort orders, and required keys. Match it exactly.
When a payload and API conflict, the API wins for record data; the payload wins for task scoping instructions.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,302 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence
- Field Conventions
- Reusable Business Rules
- Arithmetic Checks (Validate Before Output)
- Output Schema Pitfalls
- Exclusion Rules
- API Query Pattern
Files
SKILL.mdcore13,837 bytes
Read SKILL.md
# ProcureOps Task Group SKILL
## Environment
- **Base URL**: Use the URL from `environment_access.md` (`GDPEVO_ENV_BASE_URL`). This overrides any `localhost:8006` or `127.0.0.1:8006` references in task text. Do NOT start a local env or run setup scripts.
- **API endpoints**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
## Source Precedence
1. **ProcureOps API is always authoritative** for live operational records (programs, suppliers, items, contracts, requisitions, POs, receipts, invoices, payments, approvals, budgets, vendor risk).
2. **Task payloads** (memos, JSON packets) provide task-specific context — target IDs, business parameters, chargeback registers — but never override API data.
3. **Answer template** (`answer_template.json`) defines the exact output schema. Match it precisely — do not add or omit keys.
4. **as_of_date / as-of cutoff**: Filter API records to those dated on or before the task's cutoff date. Receipts, invoices, approvals, risk events — all are time-gated.
## Field ConventionsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,774 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment
- 2. Source Precedence
- 3. Field Conventions
- 4. Reusable Business Rules
- 5. Output Schema Pitfalls
- 6. Workflow SOP
- 7. Common Task Archetypes
Files
SKILL.mdcore10,690 bytesSKILL.mdcore10,690 bytes
Read SKILL.md
# ProcureOps Task Group Skill — Sourcing, Receiving, AP, and Change Control
## 1. Environment
- **API base URL**: Use the URL from `environment_access.md` (overrides any `localhost`/`127.0.0.1` in task text).
- **Public endpoints**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- Query with query parameters for filtering (e.g., `?program_id=PRG-AX17`, `?supplier_id=SUP-LUMA`).
## 2. Source Precedence
1. **ProcureOps API** — system of record for all operational data (POs, receipts, invoices, contracts, suppliers, approvals, budgets, risk events).
2. **Local chargeback register** — authoritative for chargeback statuses and amounts (train_005 pattern). When it exists, use its `status`, `reason_code`, `basis_quantity`, and `unit_cost` over API-derived values.
3. **Task payloads** (memos, packets, templates) — provide target IDs, business parameters (tax rates, as-of dates), and context; do NOT use them to override API facts.
4. When the API and a memo disagree, the API wins. When a chargeback register and the API disagree on chargeback details, the register wins.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,157 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & Sources
- Field Conventions
- Date & Time Filtering
- Arithmetic & Business Rules
- Output Schema Pitfalls
- Task-Specific Patterns
Files
SKILL.mdcore9,050 bytesSKILL.mdcore9,050 bytes
Read SKILL.md
# ProcureOps Task Group 006 — Reusable Solve Skill
## Environment & Sources
1. **API base URL**: Use the environment-provided `GDPEVO_ENV_BASE_URL`. Never use localhost unless that is the actual remote target.
2. **API is source of truth**: Task memos/payloads provide anchors (IDs, scope) and business rules. Operational data (quantities, prices, statuses, dates, risk events) comes from the API only.
3. **Endpoints used**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
4. **API data is flat**: All endpoints return full result sets — use client-side filtering, not query params.
## Field Conventions
5. **Sort ascending** for: SKU lists, ID lists (PO, receipt, invoice, supplier, risk event), nomination lines (by SKU), invoice_decisions (by invoice_id), vendor_balances (by supplier_id), program_summary (by program_id), payment hold/release queues.
6. **List fields are sets**: Unless sorting is specified in the template, the evaluator sorts. Still produce deterministic output.
7. **USD to cents**: All currency amounts rounded to 2 decimal places with standard rounding. Include `.00` suffix for whole-dollar values.
8. **Precision by field**: Ratios (receipt_completion_ratio) to 4 decimal places; percentages (quantity_variance_pct) to 1 decimal.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,946 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence
- Output Schema Rules
- Date Filtering Conventions
- Business Rules
- Arithmetic Checks
- Common Pitfalls
- Data Integrity
Files
SKILL.mdcore8,059 bytesSKILL.mdcore8,059 bytes
Read SKILL.md
# ProcureOps Task Solver — Reusable SOP
## Environment
- Use the remote API base URL provided by the task runner (env var or `environment_access.md`). Never hardcode `localhost` or `127.0.0.1`.
- Public endpoints: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_orders`, `/purchase_requisitions`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- API responses are unpaginated: `{"count": N, "results": [...]}` returns all records in one call.
## Source Precedence
1. **API is the source of truth** for all operational records. Task memos name targets but API data overrides them.
2. Local payloads (`input/payloads/`) provide business controls (tax rates, approval rules, chargeback registers) and answer templates. Use them for structure and business rules, not for record data.
## Output Schema Rules
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,775 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence
- Field Conventions
- Key Business Rules
- Arithmetic Checks
- Output Schema Pitfalls
- Exclusion Rules
- Workflow Pattern
Files
SKILL.mdcore7,481 bytesSKILL.mdcore7,481 bytesSKILL.mdcore7,481 bytes
Read SKILL.md
# ProcureOps Task Solver SOP
## Environment
- Base URL: from `environment_access.md` (GDPEVO_ENV_BASE_URL); overrides any `localhost` or `127.0.0.1` in task text.
- All data comes from the remote ProcureOps API. Do not start a local env.
- Endpoints: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- Every endpoint returns `{"count": N, "results": [...]}`. Always work from `results`.
## Source Precedence
1. **Task-local payloads** (memos, packets, chargeback registers) — provide business context, target IDs, and local-only data like chargeback registers. Treat as authoritative for business rules and target scope.
2. **API records** — source of truth for operational state (PO status, receipt quantities, invoice holds, supplier ratings, budget caps, approval events).
3. **Answer template JSON** — defines required keys, allowed values, ordering, and precision. Match it exactly.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,721 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence (critical)
- Workflow for Any Task
- Field Conventions (universal)
- Reusable Business Rules
- Exception & Blocker Codes (canonical sets)
- Output Schema Pitfalls
- Exclusion Rules
Files
SKILL.mdcore7,324 bytesSKILL.mdcore7,324 bytes
Read SKILL.md
# ProcureOps Task Solver — SOP
## Environment
- **API base URL**: `http://34.46.77.124:8006` (from environment_access.md; overrides any localhost references in task text).
- **Public endpoints**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- Never start a local env or run setup scripts. Use the remote API directly.
## Source Precedence (critical)
1. **ProcureOps API is always the system of record.** Local memos name targets and give business rules but never override live API data.
2. **Local payloads** (memos, JSON packets) provide: target IDs, business rules, date parameters, chargeback registers, release notes.
3. **answer_template.json** in each task's `input/payloads/` defines the exact output schema. Match it precisely — every key, every type, every allowed value.
## Workflow for Any Task
1. Read `input/payloads/answer_template.json` first — know the output shape before querying.
2. Read the memo/packet payload to extract: target IDs, as-of date, business rules, control parameters.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,908 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment
- 2. Source Precedence
- 3. Field & Format Conventions
- 4. Reusable Business Rules
- 5. Common Exception & Reason Codes
- 6. Blocker Codes (Nomination Readiness)
- 7. Decision Enums
- 8. Cross-Check Rules (Arithmetic Validation)
- 9. Output Schema Pitfalls
- 10. Exclusion Rules
- 11. Entity Relationship Quick Reference
- 12. Execution Checklist
Files
SKILL.mdcore11,881 bytesSKILL.mdcore11,881 bytes
Read SKILL.md
# ProcureOps Task Group — Solver SOP
## 1. Environment
- **Base URL**: `http://34.46.77.124:8006` (from `environment_access.md`). This overrides ANY `127.0.0.1:8006` or localhost reference in task text.
- **Available endpoints**: `/programs`, `/suppliers`, `/items`, `/contracts`, `/purchase_requisitions`, `/purchase_orders`, `/receipts`, `/ap/invoices`, `/ap/payments`, `/approval_events`, `/budget_snapshots`, `/vendor_risk_events`.
- **Always hit these endpoints live**; memo/packet files provide context but the API is the authoritative source of record.
## 2. Source Precedence
1. **ProcureOps API records** — authoritative for all operational data (POs, receipts, invoices, contracts, budgets, approvals, suppliers, risk events).
2. **Task-local payloads** (memos, JSON packets, chargeback registers) — provide target IDs, business rules, and contextual annotations. Use them to scope API queries but never substitute API data.
3. **Answer template JSON** — defines the exact output schema. Every field, type, precision, ordering rule, and enum constraint is mandatory.
Rule: when a memo claim and an API record conflict, the API wins.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,102 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Precedence (always follow this order)
- Public API Endpoints
- Field & Output Conventions
- Reusable Business Rules
- Exclusion Rules
- Output Schema Pitfalls
- Workflow Checklist
Files
SKILL.mdcore8,972 bytesSKILL.mdcore8,972 bytesSKILL.mdcore8,972 bytes
Read SKILL.md
# ProcureOps Task Group — Solver SOP
## Environment
- **API base URL**: `http://34.46.77.124:8006` (from `environment_access.md`).
- Always use this URL — override any `localhost`/`127.0.0.1:8006` references in task text.
- Do NOT start a local env or run setup scripts. The remote API is already live.
## Source Precedence (always follow this order)
1. **Answer template** (`answer_template.json`) — the output contract. Every required key must be present. Every enum value must match exactly. List ordering must match the template's spec (sorted-ascending vs set/unordered).
2. **ProcureOps API** — system of record for all operational entities (programs, suppliers, items, contracts, POs, requisitions, receipts, invoices, payments, approval events, budget snapshots, vendor risk events).
3. **Local memo/packet** — provides task-specific context (target IDs, business rules, close dates, chargeback excerpts, requester notes). Use these to scope your API queries but NOT as the source of truth for quantities, prices, statuses, or amounts.
## Public API EndpointsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,641 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoints (inferred from task patterns)
- Common Field Conventions
- Inventory Semantics
- Task 1: Expedite Queue (Dispatch Control)
- Task 2: Production Replenishment (BOM Kit Build)
- Task 3: Supplier Incident Scorecard
- Task 4: Allocation Desk (Transfer Wave)
- Task 5: Procurement Control (Quality Hold Review)
- Reusable Decision Precedence
Files
SKILL.mdcore15,005 bytesSKILL.mdcore15,005 bytesSKILL.mdcore15,005 bytes
Read SKILL.md
# Northwind Components ERP — Task Group 007 Skill
## Environment
**Base URL:** `http://34.46.77.124:8007`
All tasks use this shared Northwind ERP API. Ignore any prompt text that references localhost or `127.0.0.1:8007` — the remote URL above is authoritative. Do not run `env/setup.sh` or `python server.py` locally.
## API Endpoints (inferred from task patterns)
| Endpoint | Purpose |
|---|---|
| `GET /orders` | List sales orders; filter `?wave_id=...` or by order_id |
| `GET /orders/{order_id}` | Single order with lines, customer, warehouse |
| `GET /products` / `GET /products/{sku}` | Product master (name, unit_cost, supplier_id, is_active, etc.) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,190 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoints & Usage
- Inventory & Effective Availability
- Currency, Percentage & Duration Precision
- Sorting Conventions
- Task Type 1: Expedite Queue Decision
- Task Type 2: Kit Build Replenishment
- Task Type 3: Supplier Incident Scorecard
- Task Type 4: Allocation Desk / Transfer Wave
- Task Type 5: Procurement Quality Control
- General Field & Output Rules
Files
SKILL.mdcore13,455 bytesSKILL.mdcore13,455 bytes
Read SKILL.md
# Northwind ERP Operations Skill — task_group_007
## Environment
- **Base URL**: `http://34.46.77.124:8007` (from `environment_access.md`). Ignore any localhost/127.0.0.1 references in task prompts.
- All data lives behind this shared ERP API. No local .env files, no direct DB access.
## API Endpoints & Usage
The ERP exposes RESTful JSON endpoints. Known endpoints (call as needed per task):
| Endpoint | Returns |
|---|---|
| `GET /orders` | Sales orders with lines, status, warehouse, customer ref |
| `GET /products` | Product master: SKU, name, active flag, unit cost, supplier ref |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,421 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Entity ID Conventions
- Common API Endpoints
- Numeric Precision Rules
- Sorting Rules (Consistent Across All Tasks)
- Common Enum Values
- Desk 1: Expedite Queue Dispatch (train_001 pattern)
- Desk 2: Kit Build Replenishment (train_002 pattern)
- Desk 3: Supplier Incident Scorecard (train_003 pattern)
- Desk 4: Allocation Desk — Mixed-Warehouse Transfer (train_004 pattern)
- Desk 5: Procurement Quality Hold Review (train_005 pattern)
- Field Pitfalls
- Reusable SOP
Files
SKILL.mdcore17,515 bytesSKILL.mdcore17,515 bytesSKILL.mdcore17,515 bytesSKILL.mdcore17,515 bytesSKILL.mdcore17,515 bytes
Read SKILL.md
# Northwind ERP Operations — Task Group 007 Skill
## Environment
- **Base URL**: `http://34.46.77.124:8007` (always; ignore any prompt mentioning localhost or 127.0.0.1)
- All data comes from live API calls; do not use cached/static values
- Return **only** JSON matching the supplied `answer_template.json` — no narrative text outside the JSON
## Entity ID Conventions
| Entity | Pattern | Example |
|--------|---------|---------|
| Sales Orders | `SO-XXXXX` (5 digits) | `SO-70000` |
| Products / SKUs | `NW-XXXX` (4 digits) | `NW-1003` |
| Suppliers | `SUP-XXX` (3 digits) | `SUP-003` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,360 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API
- Effective Availability
- Enforced Enum Conventions
- Field Pitfalls
- Ordering Rules
- Calculations
- SOP: Expedite Queue Decision
- SOP: Kit Build Replenishment
- SOP: Allocation Desk (Mixed-Warehouse)
- SOP: Supplier Incident Scorecard
- SOP: Procurement Quality Control
Files
SKILL.mdcore10,123 bytesSKILL.mdcore10,123 bytesSKILL.mdcore10,123 bytesSKILL.mdcore10,123 bytesSKILL.mdcore10,123 bytes
Read SKILL.md
# Northwind Components ERP — Operations Skill
## API
Base URL provided at solve time. Primary endpoints:
| Endpoint | Key query params |
|---|---|
| `/orders?wave=&customer_id=&required_date=` | Wave filter, customer |
| `/orders/<order_id>` | Single order detail |
| `/products` , `/products/<sku>` | Active flag, safety_stock, overstock_threshold, weight_lb, unit_cost, supplier_id |
| `/customers` , `/customers/<customer_id>` | account_status, risk_flag, tier, margin_band |
| `/warehouses` | WH_NORTH, WH_CENTRAL, WH_WEST; zip per warehouse |
| `/inventory?warehouse_id=&sku=` | on_hand, reserved, quarantined — no `buffer` field |
| `/purchase_orders?supplier_id=&sku=&status=` | status: open, confirmed, received, cancelled; eta field |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,810 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Conventions
- Task Pattern: Expedite Queue (Dispatch Control)
- Task Pattern: Kit Build Replenishment
- Task Pattern: Supplier Incident Scorecard
- Task Pattern: Allocation Desk (Mixed-Warehouse Transfer)
- Task Pattern: Procurement Quality Control
- General Rules
Files
SKILL.mdcore12,458 bytesSKILL.mdcore12,458 bytesSKILL.mdcore12,458 bytesSKILL.mdcore12,458 bytes
Read SKILL.md
# Northwind ERP Operations Skill
## API Conventions
Base URL from `environment_access.md`; all endpoints relative to it.
### Core Endpoints
| Endpoint | Use |
|---|---|
| `/products/<sku>` | Active flag, category, supplier, safety_stock, overstock_threshold, unit_cost, weight_lb |
| `/customers/<customer_id>` | account_status, risk_flag, tier, margin_band |
| `/warehouses` | warehouse_id, zip, region |
| `/inventory?warehouse_id=&sku=` | on_hand, reserved, quarantined (omit sku for all) |
| `/orders?wave=` | Full order details with lines, warehouse, customer, shipping_speed, destination_zip |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,052 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Base
- Core Endpoints
- Inventory Effective Available
- Customer Exception Mapping
- Fulfillment Decision Logic
- Shipping Quotes
- BOM & Replenishment Calculations
- Allocation Desk (Mixed-Warehouse Transfer)
- Incident Scorecard — Recommendation Policy (Precedence)
- Incident Scorecard — Computations
- Procurement Quality Control
- Sorting & Ordering Conventions
- Currency & Precision
Files
SKILL.mdcore8,095 bytesSKILL.mdcore8,095 bytesSKILL.mdcore8,095 bytesSKILL.mdcore8,095 bytes
Read SKILL.md
# Northwind Components ERP — Replenishment & Dispatch Skill
## API Base
Use the shared ERP API base URL provided by the environment. All endpoints are read-only JSON GET.
## Core Endpoints
| Endpoint | Parameters | Notes |
|---|---|---|
| `/orders/<id>` | — | Order with lines, warehouse, shipping, customer |
| `/orders?wave=` | `wave` | All orders in a wave |
| `/products/<sku>` | — | `active`, `safety_stock`, `overstock_threshold`, `supplier_id`, `unit_cost`, `weight_lb` |
| `/customers/<id>` | — | `account_status` (active/blocked/review_required), `risk_flag` (none/fraud_watch/credit_watch), `tier` |
| `/inventory?warehouse_id=&sku=` | both required | `on_hand`, `reserved`, `quarantined` |
| `/warehouses` | — | `warehouse_id`, `zip`, `region`, `name` |
| `/shipping/quote?warehouse_id=&destination_zip=&weight_lb=&speed=` | all required | Returns `total_cost`, `service_days`, `zone_distance`. Speed: ground/two_day/overnight |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,266 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Reference
- Core Data Model Relationships
- Key Calculations
- Decision Frameworks (SOPs)
- Enum Quick Reference
- Ordering Rules (Sort Conventions)
- Common Pitfalls
Files
SKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytesSKILL.mdcore16,857 bytes
Read SKILL.md
# Northwind Components ERP — Operations Skill
Shared ERP API for dispatch, replenishment, supplier quality, allocation, and procurement control desk workflows.
## API Reference
**Base URL:** provided by the runner (see `environment_access.md` — do not assume localhost).
### Endpoints
| Endpoint | Query Parameters | Notes |
|---|---|---|
| `GET /products` | — | All products; key fields: `active`, `safety_stock`, `overstock_threshold`, `unit_cost`, `weight_lb`, `supplier_id` |
| `GET /products/<sku>` | — | Single-product lookup |
| `GET /customers` | — | All customers; `account_status` (`active`/`blocked`/`review_required`), `risk_flag` (`none`/`fraud_watch`/`credit_watch`) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,913 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Base URL
- API Endpoint Reference
- Core Calculations
- Task 1: Expedite Queue Dispatch (train_001 pattern)
- Task 2: Kit Build Replenishment (train_002 pattern)
- Task 3: Supplier Incident Scorecard (train_003 pattern)
- Task 4: Allocation Desk — Mixed Warehouse (train_004 pattern)
- Task 5: Procurement Quality Control (train_005 pattern)
- Cross-Cutting Patterns & Pitfalls
- SOP: Processing Any Train Task
Files
SKILL.mdcore24,143 bytesSKILL.mdcore24,143 bytesSKILL.mdcore24,143 bytesSKILL.mdcore24,143 bytes
Read SKILL.md
# Northwind Components ERP — Operational Skill
## Base URL
```
GDPEVO_ENV_BASE_URL=http://34.46.77.124:8007
```
All endpoints are relative to this base. No localhost, no environment setup scripts.
---
## API Endpoint Reference
| Endpoint | Method | Key Query Params | Returns |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,297 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Reference
- Key Calculation: Effective Available
- Task Pattern 1: Expedite Queue (train_001)
- Task Pattern 2: Replenishment / Kit Run (train_002)
- Task Pattern 3: Supplier Incident Scorecard (train_003)
- Task Pattern 4: Allocation Desk (train_004)
- Task Pattern 5: Procurement Quality Control (train_005)
- Cross-Cutting Rules
Files
SKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytesSKILL.mdcore20,987 bytes
Read SKILL.md
# Northwind ERP Operations Skill
## API Reference
**Base URL**: Use the shared ERP API at the URL provided by the runner. Do not use localhost.
### Endpoints
| Endpoint | Parameters | Returns |
|---|---|---|
| `GET /orders?wave=` | `wave` (wave ID) | All orders in the wave |
| `GET /orders/<order_id>` | — | Single order detail |
| `GET /products` | — | All products (54 SKUs) |
| `GET /products/<sku>` | — | Single product detail |
| `GET /customers` | — | All customers |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,335 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Step 1 — Read Input Materials
- Step 2 — API Base URL
- Step 3 — Query the Advisory API
- Step 4 — Source Resolution
- Step 5 — Tax Rates and Constants
- Step 6 — Analysis-Type-Specific Procedures
- Step 7 — Output Format Rules
- Step 8 — Common Pitfalls
- Step 9 — Quick Reference: Analysis Type to Required Sections
- Step 10 — Execution Checklist
Files
SKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytesSKILL.mdcore17,873 bytes
Read SKILL.md
# Private Wealth Advisory Structured Output Skill
## Overview
This skill produces a JSON-structured advisory output for private wealth engagements. Each task follows a consistent pipeline: read the engagement brief, query the advisory API for client records, resolve conflicting sources, compute dollar amounts with standard tax rates and formulas, select recommendation enums, and emit a JSON object conforming to a task-specific template.
## Step 1 — Read Input Materials
Three files are always present:
| File | Purpose |
|---|---|
| `input/prompt.txt` | Engagement type and client ID (always CLT-XXXX format). Names the analysis type. |
| `input/payloads/request_memo.md` | Client ID, engagement name, horizon year (if applicable), and any special instructions. |
| `input/payloads/answer_template.json` | **Authoritative schema.** Defines `required_top_level_keys`, every field path with type/enum, and any ordering rules. |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,636 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Step 1 — Identify the Engagement Type
- Step 2 — Fetch Client Data from the Advisory API
- Step 3 — Resolve Conflicting Sources
- Step 4 — Tax Policy Constants (Hardcoded Rules)
- Step 5 — Engagement-Specific Calculation Rules
- Step 6 — Output Construction Checklist
- Common Pitfalls
Files
SKILL.mdcore14,505 bytesSKILL.mdcore14,505 bytesSKILL.mdcore14,505 bytesSKILL.mdcore14,505 bytesSKILL.mdcore14,505 bytes
Read SKILL.md
# Private Wealth Advisory Structured Output Skill
## Environment
- API base URL: `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8008` (set via harness as `API_BASE`; this file overrides any localhost references).
- Use `curl` to query the advisory API. The API serves client records, source documents, account exports, life-insurance records, trust candidates, tax policy constants, and RMD factors.
- Output format: **JSON only**, conforming to `input/payloads/answer_template.json`. No prose outside the JSON object. Return valid JSON with no trailing text.
- Numbers are JSON numbers (not strings). USD amounts are rounded to cents (2 decimal places). Dates are ISO 8601 `YYYY-MM-DD`.
---
## Step 1 — Identify the Engagement Type
Read `input/payloads/answer_template.json` to determine `analysis_type` from the `fields.analysis_type` enum. Four engagement types exist:
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,228 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment
- Workflow (SOP)
- Analysis-Type Quick Reference
- Common Pitfalls
Files
SKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytesSKILL.mdcore13,504 bytes
Read SKILL.md
# Private Wealth Advisory — Task-Group Structured Output Skill
## Overview
This skill covers generating structured JSON planning outputs for private wealth advisory engagements: Roth conversion RMD analysis, ILIT Crummey funding, GRAT vs. CRAT trust comparisons, and estate liquidity action plans. The advisory API hosts client records, account exports, life-insurance records, trust candidates, tax policy constants, and RMD factors. Records may conflict across source systems — source resolution is always required.
## Environment
The remote advisory API base URL is provided by the harness. Read `environment_access.md` for the active base URL. The API is reachable over HTTP via `curl` or equivalent local command-line HTTP clients. Do **not** start a local environment or use localhost/127.0.0.1 unless the environment file itself explicitly points there.
## Workflow (SOP)
### Step 1 — Read the task inputs
Always read these files in order:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,959 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage
- Output Conventions
- Calculation Conventions
- Enum Reference
- Source Resolution Rules
- Common Pitfalls
Files
SKILL.mdcore12,357 bytesSKILL.mdcore12,357 bytesSKILL.mdcore12,357 bytesSKILL.mdcore12,357 bytesSKILL.mdcore12,357 bytesSKILL.mdcore12,357 bytes
Read SKILL.md
# Private Wealth Advisory Structured Output Skill
## Overview
This skill covers generating structured JSON planning outputs for private wealth advisory engagements. Five analysis types are supported: Roth conversion/RMD projections, ILIT Crummey funding cycles, GRAT-vs-CRAT trust comparisons, estate liquidity action plans, and near-RMD conversion comparisons. The advisory environment provides client records via REST API; source records may conflict due to imports from different advisory systems.
## API Usage
### Client Data Endpoint
- `GET /api/clients` — list all clients
- `GET /api/clients/{client_id}` — single client record
- The client object contains: `client_id`, `household_name`, `age`, `marital_status`, `filing_status`, `planning_year`, `estate_value`, `liquid_assets`, `record_status`, `advisor_team`
- `filing_status` is either `MFJ` (married filing jointly) or `SINGLE`
- `planning_year` is the current advisory year
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,115 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment Setup
- Task Input Structure
- Output Construction Rules
- Task-Specific Templates and Calculation Conventions
- Source Resolution General Rules
- Tax Constants Reference
- Common Pitfalls
- Data Derivation When API Data Is Sparse
Files
SKILL.mdcore16,481 bytesSKILL.mdcore16,481 bytesSKILL.mdcore16,481 bytesSKILL.mdcore16,481 bytesSKILL.mdcore16,481 bytes
Read SKILL.md
# Private Wealth Advisory Structured Output Skill
## Overview
This skill covers generating structured JSON planning outputs for private wealth advisory engagements. Tasks span Roth conversion and RMD analysis, ILIT Crummey funding, GRAT/CRAT trust comparisons, and estate liquidity action plans.
## Environment Setup
The advisory API provides client records at a base URL supplied by the harness. Use that base URL for all HTTP requests.
### API Endpoints
Only two endpoints are available:
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,362 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage
- Output Format Rules (All Analysis Types)
- Analysis Type Reference
- Tax & Planning Constants
- Source Resolution
- Calculation Conventions
- Common Pitfalls
- Workflow for Solving a New Task
Files
SKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytesSKILL.mdcore21,209 bytes
Read SKILL.md
# Private Wealth Advisory — Structured Planning Output Skill
## Overview
This skill covers producing structured JSON planning outputs for a private wealth advisory API. It supports four analysis types: Roth conversion with RMD projections, ILIT Crummey implementation, GRAT-vs-CRAT trust comparison, and estate liquidity action plans. The API provides client profiles; the skill specifies how to derive numeric projections, resolve conflicting source records, and conform to the answer templates.
## API Usage
### Client Lookup
```
GET {API_BASE}/api/clients → list all clients
GET {API_BASE}/api/clients/{id} → single client summary
GET {API_BASE}/api/health → service health check
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,729 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Analysis Types and Output Schemas
- General Output Conventions
- Common Pitfalls
- Workflow Recipe
Files
SKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytesSKILL.mdcore18,627 bytes
Read SKILL.md
# Private Wealth Advisory — Structured Planning Skill
## Environment
The advisory API is reachable at the base URL provided in `environment_access.md` (the `GDPEVO_ENV_BASE_URL`). All endpoints are under `/api/`. Use `curl` to fetch records; all responses are JSON arrays or objects. Query-parameter filtering is supported on list endpoints with `?client_id=CLT-XXXX`.
### Known API Endpoints
| Endpoint | Filterable | Key Fields |
|---|---|---|
| `/api/clients` | `?client_id=X` | `client_id`, `household_name`, `age`, `marital_status`, `filing_status`, `planning_year`, `estate_value`, `liquid_assets` |
| `/api/retirement-accounts` | `?client_id=X` | `account_id`, `traditional_balance`, `roth_balance`, `expected_return`, `rmd_start_age`, `recommended_conversion_years` |
| `/api/life-insurance` | no (filter client-side) | `policy_id`, `death_benefit`, `annual_premium`, `planned_contribution_date`, `is_existing_policy_transfer` |
| `/api/trust-candidates` | no (filter client-side) | `trust_case_id`, `asset_value`, `expected_growth_rate`, `grat_term_years`, `grat_annuity_rate`, `crat_term_years`, `crat_payout_rate` |
| `/api/source-documents` | `?client_id=X` | `document_id`, `source_type`, `effective_date`, `title`, `facts` (dict) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,319 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Source Document Resolution (Critical for Every Task)
- Tax Policy Constants (`GET /api/policies/tax`)
- Analysis Type 1: Roth Conversion + RMD (`roth_conversion_rmd`)
- Analysis Type 2: ILIT Crummey (`ilit_crummey_implementation`)
- Analysis Type 3: GRAT vs CRAT (`trust_comparison`)
- Analysis Type 4: Estate Liquidity Action Plan (`estate_liquidity_action_plan`)
- General Conventions
Files
SKILL.mdcore20,601 bytesSKILL.mdcore20,601 bytes
Read SKILL.md
# Private Wealth Advisory — Structured Planning Skill
## Environment
Base URL: use the `GDPEVO_ENV_BASE_URL` supplied by the harness (HTTP). All endpoints are under `/api/` and use kebab-case naming. The API is a plain JSON REST service over HTTP; no authentication is required within the staging environment. Use `curl` or equivalent. Responses are JSON arrays or objects.
### Endpoint Reference
| Endpoint | Returns | Use |
|---|---|---|
| `GET /api/clients` | Array of client objects | Look up client by `client_id` |
| `GET /api/clients/{client_id}` | Single client object | Verify or pull a specific client |
| `GET /api/retirement-accounts` | Array of IRA account objects | Get traditional/Roth balances, expected return, RMD start age, recommended conversion years |
| `GET /api/rmd-factors` | Object mapping age→factor | IRS Uniform Lifetime Table; divisor for RMD = balance / factor |
| `GET /api/life-insurance` | Array of policy objects | Death benefit, annual premium, proposed owner, contribution date, existing-policy-transfer flag |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,222 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage
- Data Model Reference
- Source Resolution Rules
- Calculation Conventions
- Common Pitfalls
- Workflow
- 2026 IRS Reference Values
Files
SKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytesSKILL.mdcore24,929 bytes
Read SKILL.md
# Private Wealth Advisory — Structured Planning Output Generator
## Overview
This skill produces structured JSON planning outputs for a private wealth advisory API. The advisory environment exposes client records, retirement accounts, life insurance policies, trust candidates, source documents with conflicting data, and IRS RMD life-expectancy factors. The task is to reconcile conflicting sources, compute tax-aware projections, and emit a JSON object that conforms to a supplied answer template.
## API Usage
### Base URL
Use the harness-supplied `API_BASE` (typically `http://<host>:8008`). Do **not** construct localhost URLs.
### Endpoints
| Method | Path | Query params | Returns |
|--------|------|-------------|---------|When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,297 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & API Base URL
- Period Convention
- Domain 1: Branch Close & Regional Reporting (Finance API)
- Domain 2: Compensation Summary (Compensation API)
- Domain 3: Compensation Forecast (Compensation API + Scenarios)
- Domain 4: Weekly Payroll Review (Payroll API)
- Rounding & Formatting Conventions
- List Ordering Conventions
- Common Pitfalls
Files
SKILL.mdcore16,675 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
## Environment & API Base URL
Use the base URL from `environment_access.md` (HTTP, not HTTPS). If a payload references `localhost` or `127.0.0.1`, override it with the remote URL from `environment_access.md`.
### Key endpoints
| Domain | Endpoint | Returns |
|--------|----------|---------|
| Finance | `/api/finance/branches` | Branch metadata (id, name, region_id) |
| Finance | `/api/finance/period-map` | M1–M24 → fiscal year mapping |
| Finance | `/api/finance/accounts` | Chart of accounts (account, category, metric_type) |
| Finance | `/api/finance/records` | Time-series values by branch × account × period |
| Compensation | `/api/compensation/rate-book` | Scale rates, seniority bands, title premiums, business rules |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,384 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- Calculation Formulas
- Output Conventions
Files
SKILL.mdcore13,558 bytes
Read SKILL.md
# Crescent Finance Ops — Transferable Skill
## Environment
All tasks use a single remote Finance Ops API. The base URL is provided in the task's `environment_access.json` (or overridden by `environment_access.md`). Always use HTTP, never HTTPS. Do not use localhost unless the override explicitly points there.
## API Reference
### Finance Domain
| Endpoint | Returns |
|---|---|
| `GET /api/finance/branches` | Array of `{branch_id, branch_name, region_id, region_name}` for all 12 branches |
| `GET /api/finance/period-map` | Array of `{fiscal_year, month_name, month_number, period}` mapping M1–M24 |
| `GET /api/finance/accounts` | Array of `{account, category, display_name, metric_type}` — currency or count |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,598 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Rounding & Output Conventions
- Task Family 1: Branch Financial Close Reporting
- Task Family 2: Compensation — Current-Year Summary
- Task Family 3: Payroll — Weekly Production Review
- Task Family 4: Regional Management Reporting
- Task Family 5: Compensation Forecast
- Common Pitfalls
Files
SKILL.mdcore17,464 bytes
Read SKILL.md
# Crescent Finance Ops — Transferable Skill
## Environment
Base URL: `http://34.46.77.124:8009` (use HTTP, never HTTPS).
This URL overrides any `localhost`/`127.0.0.1` references found in payload files.
### API Endpoints
| Endpoint | Purpose |
|---|---|
| `GET /api/manifest` | Global listing of branches, ensembles, productions, record counts |
| `GET /api/finance/branches` | Branch registry: `branch_id`, `branch_name`, `region_id`, `region_name` |
| `GET /api/finance/period-map` | Fiscal calendar: maps `period` (M1–M24) to `fiscal_year` and `month_name` |
| `GET /api/finance/accounts` | Chart of accounts: `account`, `category`, `display_name`, `metric_type` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,560 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Finance (Branch / Regional Reporting)
- Compensation (Current-Year Summary)
- Payroll (Weekly Review)
- Compensation Forecast
- Rounding Conventions
- Ordering Conventions
- Common Pitfalls
Files
SKILL.mdcore14,152 bytes
Read SKILL.md
# Crescent Finance Ops — Reusable Skill Guide
## Environment
All API calls use the environment base URL from `environment_access.md` (which overrides any localhost references in payload files). Use HTTP, not HTTPS.
Key public endpoints:
| Endpoint | Use |
|---|---|
| `GET /api/finance/branches` | Branch metadata (id, name, region_id) |
| `GET /api/finance/period-map` | Fiscal period definitions (M1–M24, month names, fiscal years) |
| `GET /api/finance/accounts` | Chart of accounts (account name, category) |
| `GET /api/finance/records` | Monthly values per account per branch; shape: `{account, branch_id, values: {M1..M24}}` |
| `GET /api/compensation/rate-book` | Pay rates, seniority bands, title premium pcts, business rules |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,838 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Domain 1: Finance Ops (Branch / Regional / Close Reporting)
- Domain 2: Compensation (Current-Year & Forecast)
- Domain 3: Payroll (Weekly Production)
- General Conventions
Files
SKILL.mdcore14,692 bytes
Read SKILL.md
# Crescent Finance Ops — Skill Reference
## Environment
All API calls use `http://34.46.77.124:8009` as the base URL. HTTP (not HTTPS).
The service exposes three domains: Finance, Compensation, and Payroll.
### Key Endpoints
| Endpoint | Purpose |
|---|---|
| GET `/api/finance/branches` | Branch registry: branch_id, branch_name, region_id |
| GET `/api/finance/period-map` | Maps period labels (M1..M24) to fiscal years and calendar months |
| GET `/api/finance/accounts` | Account definitions with categories and metric types |
| GET `/api/finance/records` | Monthly values per (branch_id, account); keyed by period M1..M24 |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,489 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Module 1 — Finance (Branch / Regional Reporting)
- Module 2 — Compensation (Current-Year Summary)
- Module 3 — Compensation (Forecast)
- Module 4 — Payroll (Weekly Production)
- Output Conventions (All Modules)
- Common Pitfalls
Files
SKILL.mdcore13,916 bytes
Read SKILL.md
# Crescent Finance Ops — Skill Reference
## Environment
All tasks target a remote Finance Ops API. Use the base URL provided in each task's `payloads/environment_access.json` (or the override in the repository's `environment_access.md`). Use HTTP, not HTTPS.
### Public API Endpoints
| Endpoint | Description |
|---|---|
| `GET /api/finance/branches` | Branch list with region assignments |
| `GET /api/finance/period-map` | Fiscal period → calendar month mapping |
| `GET /api/finance/accounts` | Chart of accounts and categories |
| `GET /api/finance/records` | All branch-period financial records |
| `GET /api/compensation/rate-book` | Pay rates, scales, premiums, quarter weeks, business rules |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,449 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Finance Ops — Branch & Regional Reporting
- Compensation Ops — Current-Year Summaries & Forecasts
- Payroll Ops — Weekly Payroll Review
- Rounding & Output Conventions (all modules)
- Common Pitfalls
Files
SKILL.mdcore16,827 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
## Environment
All API calls use HTTP against the remote base URL from `environment_access.md` / `payloads/environment_access.json`. Each task payload contains a `base_url` field that may reference localhost; **ignore it** — use the remote URL from `environment_access.md` instead.
**Remote base:** `http://<host>:<port>` (see `environment_access.md`).
Verify connectivity with `GET /health` → `{"status":"ok"}`.
---
## Finance Ops — Branch & Regional Reporting
### EndpointsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,917 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Reference
- 1. Period Convention
- 2. Income Statement Components (Finance)
- 3. Variance and Growth (Finance)
- 4. Branch and Regional Context (Finance)
- 5. Compensation — Rate Book Mechanics
- 6. Compensation — Forecast Mechanics
- 7. Payroll — Rate Book Mechanics
- 8. Payroll — Category Taxonomy
- 9. Payroll — Conflict Flags
- 10. Output Conventions
- 11. Common Pitfalls
Files
SKILL.mdcore15,612 bytesSKILL.mdcore15,612 bytes
Read SKILL.md
# Crescent Finance Ops — Transferable Skill Reference
## Environment
```
BASE = http://34.46.77.124:8009
```
Always use HTTP (not HTTPS). Call `GET /health` first to validate connectivity. The `/api/manifest` endpoint lists all available endpoints and public entity summaries (branch names, ensemble names, production titles).
## API Reference
| Domain | Endpoint | Returns |
|--------|----------|---------|
| Finance | `/api/finance/branches` | Branch list with region mappings |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,749 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Task Domains
- Rounding & Ordering Conventions
- Finance Records — Core Data Model
- Compensation — Core Data Model
- Compensation Forecast — Scenario Model
- Payroll — Core Data Model
- Common Pitfalls
Files
SKILL.mdcore14,095 bytes
Read SKILL.md
# Crescent Finance Ops — Transferable Skill
## Environment
All tasks use the same remote API base; there is no local env, no setup script, and no mock server. Always use the base URL from `environment_access.md`.
- HTTP only (no HTTPS)
- Key public endpoints:
```
GET {BASE}/api/manifest
GET {BASE}/api/finance/branches
GET {BASE}/api/finance/period-map
GET {BASE}/api/finance/accounts
GET {BASE}/api/finance/records
GET {BASE}/api/compensation/rate-bookWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,826 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment & Data Sources
- 2. Data Precedence Rule
- 3. Correlation Calculations (Pearson, Monthly Simple Returns)
- 4. Portfolio Metrics — How to Compute
- 5. Allocation Views — View, Change, Conviction
- 6. Trade Selection & Constraint Checks
- 7. Output Formatting Conventions
- 8. Workflow Checklist
- 9. Common Pitfalls
Files
SKILL.mdcore11,292 bytesSKILL.mdcore11,292 bytes
Read SKILL.md
# Asteria Investment Office — Operational Skill
## 1. Environment & Data Sources
**Base URL:** `GDPEVO_ENV_BASE_URL` (from environment_access.md). All data comes from the remote HTTP API at this URL. Never read local env source files or use localhost.
**Key endpoints:**
| Endpoint | Returns |
|---|---|
| `GET /api/catalog` | Directory of all available data |
| `GET /api/policies` | Policy thresholds (HY caps, duration bands, concentration limits) |
| `GET /api/portfolios` | List of all portfolios |
| `GET /api/portfolios/<id>` | Single portfolio: holdings, market values, current metrics |
| `GET /api/instruments/bonds` | Bond security master: YTM, modified duration, rating (IG/HY), issuer, sector/subsector, watchlist flag |
| `GET /api/issuers` | Issuer records: credit quality, watchlist status, sector |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,631 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core principle: Environment is the book of record
- API inventory (all `GET`, no auth)
- Universal conventions
- Task-type workflows
- Common pitfalls
- Quick-reference: key thresholds
- Task-agnostic output rules
Files
SKILL.mdcore14,340 bytesSKILL.mdcore14,340 bytes
Read SKILL.md
# Asteria Investment Office — Operational Skill
## Core principle: Environment is the book of record
The shared Asteria API is the **current, authoritative source**. Local payloads (desk requests, meeting memos, committee packets) are **intake context only** — they may contain stale marks, outdated worksheets, or pre-reconciled snapshots. When a local payload conflicts with the API, the API wins.
**Data-precedence value:** `"current_environment_over_stale_payload"` — use this whenever the task template includes a `data_precedence` field and the local payload carries snapshot data that the API supersedes.
## API inventory (all `GET`, no auth)
| Endpoint | Returns |
|---|---|
| `/api/catalog` | All IDs (bonds, indices, issuers, portfolios, policies, opportunity sets) |
| `/api/policies` | Constraint thresholds, allocation mapping, correlation policy, as-of date |
| `/api/portfolios/<id>` | Holdings (instrument_id, quantity_usd_m, sleeve, asset_class), market_value_usd_m, constraints, as_of_date |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,603 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Workflow: Always start by reading the API
- Task archetypes and computation rules
- Precision rules (by field type)
- Ordering rules
- Enumeration conventions
- Common pitfalls
Files
SKILL.mdcore18,405 bytesSKILL.mdcore18,405 bytes
Read SKILL.md
# Asteria Investment Office — Operational Skill
## Environment
All tasks use the shared Asteria API at `http://34.46.77.124:8010`. Treat the API as the **current book of record**. Local payload files (in `input/payloads/`) are intake context — they may contain **stale marks, outdated worksheets, or pre-refresh preferences**. When the API and a local payload disagree, the API wins.
The current environment `as_of_date` is **2026-05-29**. Always set `as_of_date` to this value in outputs. Set `data_precedence` to `"current_environment_over_stale_payload"` whenever the local payload contains stale/older data that the API supersedes.
### Key endpoints
| Endpoint | Use |
|---|---|
| `GET /api/catalog` | Inventory of all IDs (bonds, indices, issuers, portfolios, policies, opportunity sets) |
| `GET /api/portfolios` | List all portfolios with metadata, constraint policy, and current MV |
| `GET /api/portfolios/<id>` | Full portfolio detail including holdings (instrument_id, quantity_usd_m, sleeve, asset_class, notes) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,457 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Data Precedence Rule
- Key Endpoints and What They Carry
- Policy Identification
- Weighted Portfolio Metrics (Credit Tasks)
- Pearson Correlation from Index Levels (Correlation Tasks)
- View Mapping from Macro Signals (Allocation Tasks)
- Constraint Checks
- Output Ordering Conventions
- Risk Overlay Selection (Allocation Tasks)
- Precision Rules (by Field)
- Common Pitfalls
Files
SKILL.mdcore9,049 bytes
Read SKILL.md
# Asteria Investment Office — Operational Skill
## Data Precedence Rule
**Current environment always overrides stale local payloads.**
- The remote API (`/api/portfolios`, `/api/instruments/bonds`, `/api/issuers`, `/api/policies`, `/api/index-levels`, `/api/macro-signals`, `/api/allocation/prior-views`) is the book of record.
- Local request payloads contain intake context (requested ticket sizes, focus sets, review windows) but their stale snapshots, stale marks, and stale exception-board quantities must be reconciled to current API records.
- When a local payload's date or quantity conflicts with the API, use the API value.
- Set `data_precedence` fields to `"current_environment_over_stale_payload"` or equivalent.
## Key Endpoints and What They Carry
| Endpoint | Key Fields |
|---|---|
| `GET /api/catalog` | All valid IDs: bonds, indices, issuers, opportunity sets, policies, portfolios |
| `GET /api/policies` | Thresholds: `max_hy_allocation_pct` (20%), `duration_band_years` ([3.0, 5.0]), `correlation_high_threshold` (0.8), `correlation_low_threshold` (0.2), `view_score_thresholds` (OW ≥ 0.35, UW ≤ −0.35), `conviction_thresholds` (HIGH ≥ 0.7, MEDIUM ≥ 0.35, LOW < 0.35), `issuer_concentration_limit_pct` (12%), `subsector_min_count_for_diversified` (2) |
| `GET /api/portfolios/<id>` | Current holdings with `instrument_id`, `quantity_usd_m`, `sleeve`, `market_value_usd_m`, `as_of_date`, `constraint_policy_id` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,834 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment and Data Precedence
- Key API Endpoints and Their Data
- Bond Selection and Trade Construction
- Pearson Correlation from Index Levels
- Active Allocation Views
- Fixed-Income Risk Rebalancing (Rotation)
- Multi-Asset Committee Files (Cross-Task Integration)
- Common Pitfalls
Files
SKILL.mdcore15,748 bytes
Read SKILL.md
# Asteria Investment Office — Reflect Skill
## Overview
This skill covers the Asteria Investment Office workflow: energy credit trade construction, international equity correlation review, active allocation views, fixed-income risk rebalancing, and multi-asset committee decision files. All tasks share a remote HTTP API (`GDPEVO_ENV_BASE_URL`) as the authoritative book of record; local payloads are intake context only and may contain stale marks.
## Environment and Data Precedence
- The remote API at `GDPEVO_ENV_BASE_URL` is the sole current book of record. Never use localhost, `env/README.md`, or filesystem paths.
- Local payload files (desk requests, meeting memos, review packets) provide task framing but may contain stale marks. The prompt text or `environment_access.md` takes precedence over local URL references.
- **Data precedence rule**: when the API and a local payload disagree on values (quantities, dates, market values), use the API. Output `data_precedence` as `"current_environment_over_stale_payload"` whenever a stale-data warning or reconciliation note appears in the local payload. If no conflict exists, use `"no_conflict_found"`.
- The API `as_of_date` (available in `/api/policies` and each portfolio response) is the as-of date for all outputs unless a task-specific window overrides it.
## Key API Endpoints and Their Data
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,392 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core principle
- API data sources and endpoints
- Portfolio-metric calculations
- Policy thresholds
- Pearson correlation of monthly simple returns
- Bond selection rules
- Output conventions
- Common pitfalls
Files
SKILL.mdcore9,700 bytes
Read SKILL.md
# Asteria Investment Office — Operational Skill
## Core principle
Always use the **current shared Asteria environment** (remote HTTP API) as the authoritative book of record. Local payloads may carry stale marks, worksheet snapshots, or pre-reconciliation estimates. When a local payload disagrees with the API, the API wins. Express this precedence explicitly in any output field that asks for it (use `"current_environment_over_stale_payload"`).
---
## API data sources and endpoints
| Resource | Endpoint | Key fields returned |
|---|---|---|
| Catalog | `GET /api/catalog` | All ids: portfolios, bonds, indices, issuers, policies, opportunity-sets |
| Policies | `GET /api/policies` | Top-level `policy_id` (e.g., `POLICY_SET_2026_05`), allocation-mapping thresholds, credit constraints, correlation thresholds |
| Portfolios | `GET /api/portfolios/<id>` | Holdings (instrument_id, quantity_usd_m, sleeve), as_of_date, constraint_policy_id, market_value_usd_m |
| Bonds | `GET /api/instruments/bonds` | instrument_id, issuer_id, rating_bucket (IG/HY), modified_duration_years, yield_to_maturity_pct, energy_linked, candidate, subsector, recommended_theme_tags |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,439 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Workflow Rules
- 2. API/Data Usage Habits
- 3. Calculation Procedures
- 4. Output-Field Conventions
- 5. Common Pitfalls
Files
SKILL.mdcore13,482 bytes
Read SKILL.md
# Asteria Investment Office — Reusable Task-Solving Procedure
## 1. Workflow Rules
### 1.1 Always start from the API, not the local payload
Local payloads (desk requests, meeting memos, committee packets) provide **task context only** — the opportunity-set names requested, the portfolio id, the quarter, the trade size. Stale snapshots, stale marks, and stale local notes must be **overridden** with current API data.
### 1.2 Data-precedence baseline
Unless the local payload has a date *newer* than the API `as_of_date` (which it never will in this environment), the API is authoritative. Use `"data_precedence": "current_environment_over_stale_payload"` as the default answer.
### 1.3 Standard call sequence
1. `GET /api/policies` — policy thresholds, constraint bands, correlation thresholds, conviction/view-score mapping.
2. `GET /api/portfolios/<portfolio_id>` — current holdings, total market value, constraint policy id.
3. Domain-specific endpoints:
- **Credit / fixed-income:** `/api/instruments/bonds`, `/api/issuers`, `/api/market/energy`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,152 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment & Data Precedence
- 2. Policy Lookup Chain
- 3. Task-Type Procedures
- 4. Common Formulas
- 5. Output Conventions
- 6. Common Pitfalls
Files
SKILL.mdcore20,577 bytes
Read SKILL.md
# Asteria Investment Office — Reusable Task-Solving Procedure
This skill covers institutional portfolio risk tasks in the Asteria Investment Office domain: credit trade construction, equity correlation reviews, active allocation view refreshes, fixed-income risk rebalancing, and multi-asset committee decision files.
---
## 1. Environment & Data Precedence
**API base URL** as provided by the task environment. All data comes from these HTTP GET endpoints:
| Endpoint | What it returns |
|---|---|
| `/api/catalog` | All valid IDs: portfolios, instruments, issuers, indices, policies, opportunity sets |
| `/api/policies` | Constraint thresholds, allocation mapping rules, correlation thresholds, all policy detail |
| `/api/portfolios` | Portfolio summaries (market value, policy link, strategy) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,199 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & Data Precedence
- API Data Model Quick Reference
- Task-Type Workflows
- Precision & Rounding Rules
- Sorting & Ordering Conventions
- Common Pitfalls
- Output Conventions
Files
SKILL.mdcore16,674 bytesSKILL.mdcore16,674 bytes
Read SKILL.md
# Asteria Investment Office — Task-Solving Skill
## Environment & Data Precedence
**Single source of truth**: The remote Asteria API at `GDPEVO_ENV_BASE_URL` (set per session). All portfolio, bond, issuer, index, policy, allocation, and signal data lives there. The root `GET /` enumerates available endpoints.
**Stale-data rule**: Every local payload (desk requests, memos, committee packets, stale snapshots) may contain outdated marks, quantities, or preferences. The API environment is the current book of record. When local data conflicts with API data, the API wins. The `data_precedence` field in relevant templates should be `"current_environment_over_stale_payload"` whenever the API is used to resolve or override local numbers.
**as_of_date**: Always pulled from the current portfolio record on the API (`/api/portfolios/<id>` → `as_of_date`). Across the 2026-05 environment this is `"2026-05-29"`. Use it for every output's `as_of_date` field.
---
## API Data Model Quick Reference
| Endpoint | Key fields | Used by tasks |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,440 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoints
- Risk Rating Derivation (Dominant-Factor Rule)
- CDFI Factor Scoring
- Stress Testing
- CRE Weighted Score (Competing CRE Decisions)
- Concentration Calculations
- NPA Benchmark Analysis
- Watch-List Action Assignment
- Application Decision Rules
- Material Downgrades
- Top Problem Credit
- CU Segment Posture
- Sorting Conventions
- Numeric Precision Rules
- Common Pitfalls
Files
SKILL.mdcore16,389 bytes
Read SKILL.md
# Credit Office Lending Committee Skill
## Environment
- Base URL: `http://34.46.77.124:8011`
- Start with `GET /api/manifest` for endpoint inventory and `GET /api/health` for record counts.
- Benchmark versions: FDIC `fdic_q4_2024`, NCUA `ncua_q1_2025`. Policy version: `credit_policy_v2025Q1`.
## API Endpoints
| Endpoint | Use |
|---|---|
| `GET /api/manifest` | Endpoint inventory, benchmark versions, record counts |
| `GET /api/health` | Service status, record counts |
| `GET /api/branches` | All branches (id, name, lending_capacity_q1, sector_ceiling_pct, cre_policy_limit_pct, institution_type, state_code, total_assets) |
| `GET /api/branches/{id}` | Single branch details |
| `GET /api/branches/{id}/metrics` | Quarterly metrics (nonperforming_loans, total_loans_outstanding, delinquency_30_plus_pct, allowance, net_charge_offs) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,371 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Workflow
- Numeric Precision (Universal)
- Sorting Rules (Universal)
- Common Enum Sets
- Task-Specific Business Rules
- Common Pitfalls
Files
SKILL.mdcore13,243 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending Committee — SKILL
## Environment
Base URL: `http://34.46.77.124:8011`. Never use `localhost`/`127.0.0.1` or run `env/setup.sh`.
## API Workflow
1. **Discover**: `GET /api/manifest` → `GET /api/health`
2. **Branch data**: `GET /api/branches`, `GET /api/branches/{id}`, `GET /api/branches/{id}/metrics`, `GET /api/branches/{id}/loans`, `GET /api/branches/{id}/sector-exposures`
3. **Applications**: `GET /api/applications` (filter by branch)
4. **Policies & benchmarks**: `GET /api/policies`, `GET /api/benchmarks` (or benchmark-specific endpoints from manifest)
5. **Segment**: `GET /api/segments/{segment_id}` and related NCUA benchmark endpoints
## Numeric Precision (Universal)When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,659 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Workflow
- Business Rules
- Output Conventions
- Common Pitfalls
Files
SKILL.mdcore13,335 bytes
Read SKILL.md
# Credit Office Lending Committee Skill
## Environment
- **Base URL**: `http://34.46.77.124:8011` (never use localhost)
- **Policy version**: `credit_policy_v2025Q1`
- **Benchmarks**: FDIC Q4 2024 (`fdic_q4_2024`), NCUA Q1 2025 (`ncua_q1_2025`)
## API Workflow
Always start with `GET /api/manifest` to confirm endpoints, then fetch data in parallel where possible:
| Endpoint | Returns |
|---|---|
| `GET /api/health` | Service status, record counts |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,017 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Workflow
- Risk-Rating Rules
- CDFI Factor Scoring (for Risk Classes)
- CRE Weighted Score
- Stress-Test Conventions
- Concentration Rules
- Numeric Precision
- Sorting Conventions
- Key Enum Values
- Common Pitfalls
Files
SKILL.mdcore7,524 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending-Committee Skill
## API Workflow
1. Start with `GET /api/manifest` to discover endpoints, benchmark versions, and record counts.
2. `GET /api/policies` for business rules: risk-rating thresholds, CDFI factor scores, CRE scoring weights, stress formulas, concentration policies.
3. Fetch target-branch data: `GET /api/branches/{id}` (metadata, limits), `GET /api/branches/{id}/metrics` (financials — use most recent quarter), `GET /api/branches/{id}/loans` (portfolio), `GET /api/branches/{id}/sector-exposures` (concentrations), `GET /api/branches/{id}/applications` (pending apps).
4. Benchmarks: `GET /api/benchmarks/fdic/q4-2024` and `GET /api/benchmarks/ncua/q1-2025`.
5. Credit-union segments: `GET /api/credit-union-segments/{id}`.
6. Use the base URL from `environment_access.md`; never hardcode localhost.
## Risk-Rating Rules
**Dominant-factor rule:** Final re-derived rating = **worst** (maximum numeric) rating from *available* DSCR, LTV/collateral, and delinquency factors. Skip null factors.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,135 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Workflow
- Risk Rating Re-derivation (Dominant Factor Rule)
- CDFI Factor Scoring
- Stress-Test Formulas
- CRE Weighted Scoring (Competing Credit)
- Concentration Rules
- Benchmark Conventions
- Output Field Conventions
- Common Pitfalls
Files
SKILL.mdcore7,789 bytes
Read SKILL.md
# Credit Risk Lending Committee Skill
## API Workflow
Base URL from environment (not localhost). Start with:
- `GET /api/manifest` — lists endpoints, benchmark versions, record counts
- `GET /api/health` — confirms service status
- `GET /api/policies` — risk-rating thresholds, CDFI factor tables, stress formulas, CRE weights, concentration rules
Key data endpoints:
- `GET /api/branches` / `GET /api/branches/{id}` — branch metadata, lending capacity, sector ceiling, CRE limit, state, institution type
- `GET /api/branches/{id}/metrics` — quarterly metrics (NPA, total loans, delinquency, deposits)
- `GET /api/branches/{id}/loans` — loan portfolio with ratings, DSCR, LTV, payment status, FICO, debt-to-asset, liquidity
- `GET /api/branches/{id}/sector-exposures` — per-sector current exposure, limit_pct, grandfathered flag
- `GET /api/branches/{id}/applications` — pending applications with DSCR, LTV, FICO, DTI, guarantor, relationship, SBA, bankruptcyWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,017 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Usage Workflow
- Risk Rating Re-derivation (Dominant Factor Rule)
- CDFI Factor Scoring
- Watch-List Action Mapping
- DSCR Stress Tests
- CRE Weighted Score
- Concentration and Sector Limits
- FDIC Benchmark Comparisons
- NCUA Benchmark Comparisons
- Output Field Conventions
- Common Pitfalls
Files
SKILL.mdcore11,844 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending Committee — Skill Reference
## API Usage Workflow
1. **GET /api/manifest** — discover available endpoints, benchmark versions, and record counts.
2. **GET /api/branches/{branch_id}** — branch metadata: lending capacity, policy limits, sector ceilings, state, institution type.
3. **GET /api/branches/{branch_id}/metrics** — quarterly metrics: total loans outstanding, nonperforming loans, delinquency pct, deposits, ALLL, net charge-offs. Use the most recent quarter (typically 2025Q1) unless the task specifies otherwise.
4. **GET /api/branches/{branch_id}/loans** — full loan portfolio with ratings, DSCR, LTV, payment status, CDFI factors, exposure, sector, borrower details.
5. **GET /api/branches/{branch_id}/sector-exposures** — per-sector current exposure, limit_pct, and grandfathered flag.
6. **GET /api/branches/{branch_id}/applications** — pending applications with credit metrics, sector, and requested amounts.
7. **GET /api/policies** — risk-rating thresholds, CDFI factor scores, CRE weighted-score weights, DSCR stress formulas, concentration rules, action enums.
8. **GET /api/benchmarks/fdic/q4-2024** — FDIC benchmark ratios (total_loans_noncurrent_pct, total_real_estate_noncurrent_pct, total_real_estate_30_89_pct, construction_development_noncurrent_pct).
9. **GET /api/benchmarks/ncua/q1-2025** — state-level NCUA metrics: delinquency_bps, loan_to_share_pct, roaa_bps, positive_net_income_pct.
10. **GET /api/credit-union-segments/{segment_id}** — segment details: capacity, peer states, minimum checklist, risk tolerance, internal context.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,034 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Usage
- Risk Rating Re-derivation (Dominant Factor Rule)
- CDFI Factor Scoring (for watch-list / workout tasks)
- CRE Weighted Scoring (for competing CRE decisions)
- Stress Testing
- NPA Benchmark Variance (FDIC)
- Concentration Rules
- Application Decision Logic
- Credit Union Segment Posture
- Watch-List Actions Enum
- Sorting / Ordering Conventions (CRITICAL — the templates enforce these)
- Rounding / Precision Rules
- Common Pitfalls
Files
SKILL.mdcore14,952 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending Committee Skill
## API Usage
- **Base URL**: `http://34.46.77.124:8011` (from `environment_access.md` — ALWAYS override localhost references)
- **Start**: `GET /api/manifest` and `GET /api/health` to confirm connectivity and see record counts
- **Key endpoints**:
- `GET /api/branches` — all branches (id, state, ceilings, capacity)
- `GET /api/branches/{id}` — single branch detail
- `GET /api/branches/{id}/metrics` — quarterly metrics (NPA, deposits, delinquency, charge-offs)
- `GET /api/branches/{id}/loans` — loan portfolio with ratings, DSCR, LTV, FICO, payment status
- `GET /api/branches/{id}/sector-exposures` — per-sector exposure, limit_pct, grandfathered flag
- `GET /api/branches/{id}/applications` — pending applications
- `GET /api/policies` — credit policy (ratings, CDFI, CRE score, stress, concentration rules)
- `GET /api/benchmarks/fdic/q4-2024` — FDIC benchmark ratiosWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,016 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Workflow
- Risk Rating Derivation (Dominant Factor Rule)
- CDFI Factor Scoring (for watch-list / adverse-rated loans)
- CRE Weighted Score (for CRE applications)
- Stress Tests
- Concentration Rules
- Application Decisioning
- Credit Union Segment Posture
- Benchmark Comparison
- Output Field Conventions
- Common Pitfalls
Files
SKILL.mdcore10,923 bytesSKILL.mdcore10,923 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending-Committee Skill
## API Workflow
**Base URL**: `http://34.46.77.124:8011` (override any localhost/setup.sh references).
**Discovery**: Start with `GET /api/manifest` (lists endpoints, benchmark versions, counts). Then `GET /api/health`.
**Key endpoints** (see manifest for all):
| Endpoint | Use |
|---|---|
| `/api/branches` | Branch list with `lending_capacity_q1`, `sector_ceiling_pct`, `cre_policy_limit_pct`, `total_assets`, `institution_type`, `fdic_benchmark_set`, `state_code` |
| `/api/branches/{id}/metrics` | Quarterly metrics: `nonperforming_loans`, `total_loans_outstanding`, `delinquency_30_plus_pct`, `total_deposits`, `allowance_for_loan_losses`, `net_charge_offs`. Use latest (`2025Q1`). |
| `/api/branches/{id}/loans` | Full loan book: ratings, DSCR, LTV, payment_status, collateral_value, sectors, etc. |
| `/api/branches/{id}/sector-exposures` | Per-sector `current_exposure`, `limit_pct`, `grandfathered` flag |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,322 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & API Entry
- Risk Rating Re-Derivation (Dominant Factor Rule)
- CDFI Risk-Class Scoring
- CRE Weighted Credit Scoring
- Stress-Test Formulas
- Concentration Rules
- NPA Benchmark Calculation
- Credit Union Segment Posture
- Allocation & Decision Logic
- Output Conventions (MUST FOLLOW)
- Common Pitfalls
Files
SKILL.mdcore11,881 bytes
Read SKILL.md
# Bank Branch Credit-Risk Lending-Committee Skill
## Environment & API Entry
Base URL: `http://34.46.77.124:8011` (from `environment_access.md`; overrides any localhost refs in task prompts).
Start every session with `GET /api/manifest` to confirm benchmark versions and available endpoints.
### Core Endpoints
| Endpoint | Returns |
|---|---|
| `GET /api/manifest` | Benchmark versions, record counts, endpoint list |
| `GET /api/branches` | All branches with `branch_id`, `institution_type`, `lending_capacity_q1`, `sector_ceiling_pct`, `cre_policy_limit_pct`, `state_code`, `total_assets` |
| `GET /api/branches/{id}` | Single branch detail |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,738 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoint Inventory & Data Model
- Status Precedence (Universal Rule)
- Source Precedence (Conflict Resolution)
- Five Task Patterns
- Sorting & Selection Rules
- Audit Scope Isolation (Critical)
- Common Pitfalls
- Output Conventions
Files
SKILL.mdcore19,819 bytesSKILL.mdcore19,819 bytesSKILL.mdcore19,819 bytesSKILL.mdcore19,819 bytes
Read SKILL.md
# People Lifecycle HRMS — Employee Lifecycle & Policy Operations SOP
## Environment
All API calls go to `http://34.46.77.124:8012`. There is no localhost or port
variable — ignore any task text that references localhost, 127.0.0.1, or
`<port>`. The login credential `ops.lead@peopleops.local / PeopleOps#2026` is
the single ops operator identity used across all tasks.
## API Endpoint Inventory & Data Model
| Endpoint | Returns | Key fields |
|---|---|---|
| `GET /api/summary` | Counts, departments, case breakdown | `departments[].department_id`, `counts` |
| `GET /api/employees` | All 44 employees | `employee_id`, `name`, `department`, `leave_balance_days`, `status`, `hire_date`, `salary_band`, `manager` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,979 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- API Endpoint Reference
- Business Domain Workflows
- Cross-Cutting Rules
- Common Pitfalls
Files
SKILL.mdcore20,916 bytesSKILL.mdcore20,916 bytesSKILL.mdcore20,916 bytesSKILL.mdcore20,916 bytes
Read SKILL.md
# People Lifecycle HRMS — Operations Skill
## Environment
- **Base URL**: `http://34.46.77.124:8012`
- **Credentials**: `ops.lead@peopleops.local` / `PeopleOps#2026`
- All endpoint references below are relative to the base URL.
## API Endpoint Reference
| Endpoint | Purpose |
|---|---|
| `GET /api/manifest` | Module inventory, file counts, entry points |
| `GET /api/summary` | Case counts by status, department list |
| `GET /api/employees` | All employee profiles (44 total) |
| `GET /api/cases` | All cases with status, owner, policy refs |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,313 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Endpoints
- General API Workflow
- Core Business Rules
- Task-Type Specific SOPs
- Audit Event Handling
- Enum Reference
- Common Pitfalls
Files
SKILL.mdcore18,675 bytesSKILL.mdcore18,675 bytesSKILL.mdcore18,675 bytesSKILL.mdcore18,675 bytesSKILL.mdcore18,675 bytes
Read SKILL.md
# People Lifecycle HRMS — Employee Lifecycle & Policy Operations
## Overview
This skill covers the ERP HR employee-lifecycle and policy-operations task group:
onboarding closeout, leave source precedence, payroll assignment readiness,
recruitment reconciliation, and policy case folder/notice review. All work is
done against a shared remote People Lifecycle HRMS API at the base URL provided
in `environment_access.md`.
## API Endpoints
| Endpoint | Returns |
|---|---|
| `GET /api/manifest` | Business modules, entry points, file counts, seed, generation date |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,634 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Workflow
- Record-Type Discrimination in Payroll Ledgers
- Source Precedence Rules
- Audit Event Handling
- Case Approvals
- Recruitment Reconciliation
- Output Field Conventions
- Date and Sorting Rules
- Common Pitfalls
Files
SKILL.mdcore15,652 bytes
Read SKILL.md
# ERP HR Employee Lifecycle & Policy Operations
## Overview
This skill covers PeopleOps verification and reconciliation workflows across
employee onboarding, leave management, payroll assignments, policy-case folder
readiness, formal-notice quality, recruitment-pipeline reconciliation, and
audit-driven control decisions. All workflows operate against the shared
People Lifecycle HRMS API.
## API Workflow
### Key Endpoints
| Endpoint | Purpose |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,698 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Workflow
- Business Rules
- Output Field Conventions
- Date, Calculation, and Sorting Rules
- Source Precedence (Cross-Domain)
- Common Pitfalls
Files
SKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytesSKILL.mdcore11,975 bytes
Read SKILL.md
# ERP HR Employee-Lifecycle & Policy-Operations Skill
## Overview
This skill covers verification and reconciliation tasks across the People Lifecycle
HRMS: onboarding closeout, remote-work policy case review, recruitment
reconciliation, leave source precedence, and payroll assignment readiness.
## API Usage Workflow
### Step 1 — Orient
Fetch the manifest (`GET /api/manifest`) for available modules and record counts,
then the summary (`GET /api/summary`) for case statuses and department listings.
### Step 2 — Gather evidence by domainWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,980 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Workflow
- Core Business Rules
- Output Field Conventions
- Sorting Rules
- Common Pitfalls
Files
SKILL.mdcore13,200 bytesSKILL.mdcore13,200 bytesSKILL.mdcore13,200 bytes
Read SKILL.md
# ERP HR Employee-Lifecycle & Policy-Operations Skill
## Overview
This skill covers PeopleOps verification tasks across the employee lifecycle:
onboarding closeout, leave source precedence, payroll assignment readiness, policy
case folder/notice review, and recruitment reconciliation. The system provides
REST APIs for employees, payroll ledgers, cases, policies, documents, messages,
audit events, and recruitment data.
## API Usage Workflow
The `environment_access.md` entrypoint provides the base URL. All data is
read-only via GET endpoints. The canonical workflow for any lifecycle
verification follows the evidence hierarchy below.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,401 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Workflow
- Core Business Rules
- Record Statuses and Their Meaning
- Output Field Conventions
- Calculation Rules
- Date and Sorting Rules
- Evidence Source Order
- Common Pitfalls
Files
SKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytesSKILL.mdcore15,246 bytes
Read SKILL.md
# ERP HR Employee-Lifecycle and Policy-Operations Skill
## Overview
This skill covers five related HRMS operations:
1. **Onboarding closeout** — verifying leave and payroll setup before approving close
2. **Remote-work policy case review** — folder readiness and formal notice quality
3. **Recruitment reconciliation** — candidate outcomes, offer verification, notice follow-up, and payroll handoff
4. **Leave source precedence** — determining the authoritative leave policy when sources conflict
5. **Payroll assignment and accrual readiness** — verifying submitted salary assignment and accrual batch status
---
## API Usage Workflow
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,917 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Workflow
- Business Rules by Module
- Audit Event Model
- Output Field Conventions
- Date, Calculation, and Sorting Rules
- Source Precedence (Cross-Module)
- Common Pitfalls
Files
SKILL.mdcore21,368 bytesSKILL.mdcore21,368 bytesSKILL.mdcore21,368 bytesSKILL.mdcore21,368 bytesSKILL.mdcore21,368 bytesSKILL.mdcore21,368 bytes
Read SKILL.md
# ERP HR Employee Lifecycle & Policy Operations Skill
## Overview
This skill covers People Operations tasks in the Northwind People HRMS: onboarding closeout,
leave source precedence, policy case review (folder readiness + formal notice quality),
recruitment reconciliation, and payroll assignment/accrual readiness. The system exposes a
REST API and a set of interrelated records — employee profiles, payroll ledgers, cases,
policies, documents, messages, audits, and recruitment packets. Decisions always flow from
**authoritative records** (approved/submitted ledger assignments) over summary snapshots
(employee profiles, case summaries, messages alone).
---
## API Usage WorkflowWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,163 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Reference
- Core Business Rules
- Audit Event Scoping
- Recruitment Reconciliation Workflow
- Closeout and Remediation
- Field Reference — Normalized Enum Values
- Data Flow: Step-by-Step Workflow for Each Task Type
- Common Pitfalls
- Sorting Rules
- Date and Period Conventions
- Policy Reference Quick Card
Files
SKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytesSKILL.mdcore25,887 bytes
Read SKILL.md
# ERP HR Employee Lifecycle & Policy Operations — Skill Guide
## Overview
This skill covers PeopleOps lifecycle closeout verification, leave-source precedence,
policy-case folder/notice review, recruitment reconciliation, and payroll-assignment
readiness checks against the Northwind People HRMS.
**Environment:** Use the API at `GDPEVO_ENV_BASE_URL` (do not use localhost). All data
is read-only via GET endpoints.
---
## API Reference
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,927 tokensWhen the skill is activated
Reference documents
4,197 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- When to use this skill
- Step 1 — Discover the base URL
- Step 2 — Read the task prompt and answer template
- Step 3 — Identify the task type
- Step 4 — Gather data from the API
- Step 5 — Cross-reference and apply business rules
- Step 6 — Build the JSON response
- Step 7 — Validate before returning
- Supporting files
Files
SKILL.mdcore8,677 bytesSKILL.mdcore8,677 bytesapi_reference.mddoc5,125 bytesbusiness_rules.mddoc11,635 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Reusable Skill
## Overview
This skill handles healthcare intake coordination tasks against the Cedar Ridge Intake Coordination Portal, a REST API serving patient, referral, transfer, document, chart, pharmacy, ICD, and program-candidate data. Tasks arrive as a natural-language prompt paired with a JSON answer template that specifies the exact output shape.
## When to use this skill
Invoke this skill whenever a task:
- Mentions the Cedar Ridge Intake Coordination Portal or a `<TASK_ENV_BASE_URL>` placeholder.
- Requires building a structured JSON response from an answer template.
- Involves healthcare intake workflows: patient access verification, referral auditing, transfer review, chronic-care enrollment panels, or chart-activation reconciliation.
## Step 1 — Discover the base URL
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,561 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- When to Use
- Step 1 — Resolve the Environment
- Step 2 — Read the Task Prompt
- Step 3 — Read and Internalize the Answer Template
- Step 4 — Gather Data from the Portal
- Step 5 — Apply Business Rules and Classify
- Step 6 — Assemble the Output JSON
- Step 7 — Validate and Return
- Common Pitfalls
Files
SKILL.mdcore11,649 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination Portal — Reusable Task Skill
## Overview
This skill completes a healthcare intake-coordination task by querying the Cedar Ridge Intake Coordination Portal REST API and producing a single JSON response that follows a provided answer template. The tasks span domains such as patient access verification, referral auditing, transfer review, enrollment paneling, and referral-to-chart activation, but all follow the same core workflow.
## When to Use
Invoke this skill whenever the task prompt references:
- The "Cedar Ridge Intake Coordination Portal" (or similar healthcare intake system)
- A `<TASK_ENV_BASE_URL>` placeholder that must be resolved to a real base URL
- An `input/payloads/answer_template.json` file that defines the required output shape
- A batch, roster, program code, or other grouping identifier whose data lives behind the portal
## Step 1 — Resolve the EnvironmentWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,177 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Step 1 — Parse the task prompt
- Step 2 — Understand the API surface
- Step 3 — Fetch data from the portal
- Step 4 — Read and internalize the answer template
- Step 5 — Cross-reference and analyze
- Step 6 — Build the output JSON
- Step 7 — Validate before returning
Files
SKILL.mdcore9,570 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Agent Skill
## Overview
This skill covers healthcare-intake audit, verification, and enrollment-panel tasks that run against the **Cedar Ridge Intake Coordination Portal** — a REST API serving patient, referral, transfer, document, chart, ICD, pharmacy, program-candidate, and SQL-query endpoints.
When you receive a task that references this portal, follow the workflow below. It is designed to be reusable across different batch types (new-patient rosters, referral audits, transfer reviews, chronic-care enrollment panels, and referral-to-chart activation).
---
## Step 1 — Parse the task prompt
Read the provided `prompt.txt` (or equivalent task description) and identify:
| Signal | What to extract |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,981 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Core Workflow
- Task-Type Patterns
- Common Pitfalls
Files
SKILL.mdcore9,264 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination Portal — Solver Skill
## Overview
This skill covers how to solve structured healthcare intake tasks against the Cedar Ridge Intake Coordination Portal. The portal exposes read-only REST endpoints for patients, referrals, transfers, documents, charts, programs, ICD codes, pharmacies, and a read-only SQL query interface.
## Core Workflow
### 1. Read the prompt and payloads first
- Open the task `prompt.txt` to identify the target batch/roster/program and the task type.
- Open `payloads/answer_template.json` to understand the **exact required JSON shape** before querying any data.
- The template defines required keys, allowed enum values, list ordering, and numeric precision. Deviating from these constraints produces wrong answers.
### 2. Discover and query the APIWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,724 tokensWhen the skill is activated
Reference documents
997 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Portal Data Model
- Systematic Approach
Files
SKILL.mdcore12,172 bytesSKILL.mdcore12,172 bytesdata_model.mddoc4,120 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination Portal — Skill
## Overview
This skill covers intake coordination tasks for the Cedar Ridge healthcare system. The portal exposes patient, referral, transfer, chart, program, and clinical data through REST endpoints and a read-only SQL interface. Every task follows a consistent pattern: gather data from the portal, cross-reference entities, apply decision rules derived from the answer template's controlled vocabularies, and produce a structured JSON response with patient-level detail and cohort-level summary counts.
## Portal Data Model
All data lives in a SQLite-backed system. Use the read-only `POST /query` endpoint with standard SQL to fetch data efficiently across entities. The GET endpoints return joined views (e.g., `/patients/{id}` returns patient demographics, coverage, PBM, pharmacy assignments, chart artifacts, and rosters in one call). Prefer SQL for bulk cross-entity queries; use GET endpoints to verify single-entity views or when the join logic is unclear.
### Core Tables
| Table | Key columns | Notes |
|---|---|---|
| `patients` | patient_id, existing_chart, preferred_contact, phone, email, address, emergency_contact_present | Demographics and contact preferences |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,101 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Workflow
- Common pitfalls
Files
SKILL.mdcore9,562 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Task Solver
## Overview
This skill covers solving structured data-reconciliation tasks against the Cedar Ridge Intake Coordination Portal REST API. Each task provides a prompt describing a clinical-intake scenario, an answer template specifying the exact JSON output shape, and access to a shared read-only API with patient, referral, transfer, chart, document, and program data.
## Workflow
### 1. Read all inputs first
Before making any API call, read every file in the task's `input/` directory:
- **`prompt.txt`** — The task description. Identifies the target entity (roster ID, batch ID, transfer batch, program code) and the clinical scenario. Note any overrides for the base URL.
- **`payloads/answer_template.json`** — The required JSON output shape. This is your contract. Study every required key, every allowed enum value, every ordering constraint, and every data type before you write a single line of output.
- **`payloads/target_roster.json`** or similar — If present, lists the specific entity IDs the task operates on.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,845 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use this skill
- Portal overview
- Operating rules (distilled from train evidence)
- Task-type patterns
Files
SKILL.mdcore8,548 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Reusable Operating Rules
## When to use this skill
Invoke this skill whenever the task involves the **Cedar Ridge Intake Coordination Portal** — a healthcare intake system that exposes patient, referral, transfer, chart, document, ICD, pharmacy, and program-candidate records through a REST API plus a read-only SQL endpoint. The skill applies to any intake workflow: patient access verification, referral audit, transfer review, chronic-care enrollment, or referral-to-chart activation.
## Portal overview
All endpoints live under the base URL provided as `<TASK_ENV_BASE_URL>`. Credentials are not required.
### Available endpoints
| Method | Path | Purpose |
|--------|------|---------|
| GET | `/` | Health / connectivity check |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,019 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Phase 1 — Orient
- Phase 2 — Gather
- Phase 3 — Reconcile
- Phase 4 — Assemble
- Phase 5 — Validate
- Common pitfalls
- Supporting files
Files
SKILL.mdcore13,142 bytesSKILL.mdcore13,142 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Agent Operating Procedure
## Purpose
This skill covers the reusable workflow for completing Cedar Ridge healthcare intake coordination tasks. Each task instance provides a natural-language prompt, an answer template JSON schema, and optionally additional input payloads. The agent queries a REST + SQL API to gather records, reconciles data across endpoints, applies business rules encoded in the template's controlled vocabularies, and returns a single JSON object conforming exactly to the template.
Do **not** hard-code task-specific values (roster IDs, batch IDs, patient lists, program codes, dates). Read them from the prompt and template. All tasks follow the same skeleton; this document describes the skeleton.
---
## Phase 1 — Orient
### 1.1 Read the environment manifest
Read `environment_access.md` (project root, or at the path noted in the prompt). It supplies:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,233 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Pre-Flight
- Portal Interaction Pattern
- Domain-Specific Patterns
- Controlled Vocabulary Reference
- Output Formatting Rules
- Common Pitfalls
- Task-Type Quick Dispatch
Files
SKILL.mdcore18,326 bytes
Read SKILL.md
# Cedar Ridge Intake Coordination — Reusable Operating Rules
## Purpose
Execute intake-coordination tasks against the Cedar Ridge Intake Coordination Portal. This skill covers patient access verification, referral auditing, transfer review, chronic-care enrollment panel assembly, and referral-to-chart activation. Adapt the workflow to whichever intake domain the prompt specifies.
## Pre-Flight
### Environment
The portal base URL and allowed endpoints are defined in `environment_access.md` at the repository root. Read that file first and resolve `<TASK_ENV_BASE_URL>` (or `TASK_ENV_BASE_URL`) placeholders in prompts to the `base_url` value it declares. Use only the endpoints listed in its `allowed_endpoints` block; do not invent URLs.
```
allowed route set (subject to change — always consult environment_access.md):
GET /When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,336 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Task Inputs
- Environment Access
- Evidence Gathering
- Criteria Evaluation
- Source Precedence
- Constructing the Basis Audit
- Answer Construction
- Task Execution Flow
- Domain Notes
Files
SKILL.mdcore11,457 bytesSKILL.mdcore11,457 bytesSKILL.mdcore11,457 bytes
Read SKILL.md
# Northstar Health Plan Payer Operations Review
Perform structured business reviews for Northstar Health Plan payer operations — utilization management, appeals, payment integrity, peer-to-peer, and margin analysis — by querying the shared payer environment and returning a determination that conforms to the supplied answer template.
## Task Inputs
Every task provides three inputs in the working directory under `input/`:
| File | Purpose |
|---|---|
| `input/prompt.txt` | Natural-language task description identifying the target business object, requester role, reporting date, and what to produce. |
| `input/payloads/task_context.json` | Structured metadata: task_id, target business identifier, requester role, environment configuration block, and local memos with domain-specific guidance. |
| `input/payloads/answer_template.json` | Exact required JSON shape. Lists `required_top_level_keys`, field types, enum choices, ordering rules, numeric precision, and date format. |
The answer template is the contract. Every required key must appear in the output. Enum values must match the listed choices exactly. Ordering rules on lists must be followed. The `basis_audit` sub-object is required in every Northstar determination.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,916 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Input Files
- Environment Access
- Workflow
- Common Enumeration Tables
- Cross-Cutting Rules
- Step-by-Step Execution Template
Files
SKILL.mdcore18,388 bytesSKILL.mdcore18,388 bytesSKILL.mdcore18,388 bytes
Read SKILL.md
# Northstar Health Plan Payer Operations Skill
## Purpose
Execute structured payer-operations tasks for Northstar Health Plan using the shared
payer-operations environment. Each task requires reading a task context, gathering
records from the environment, applying business rules, and returning a single JSON
object that conforms to a supplied answer template.
## When to Use
Invoke this skill when the task involves:
- A Northstar Health Plan business identifier (case, appeal, claim, queue, P2P event)
- A shared payer-operations environment accessed over HTTP
- A structured JSON output conforming to an `answer_template.json` schemaWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,649 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Read the Inputs
- 2. Access the Environment
- 3. Collect and Cross-Reference Evidence
- 4. Construct the Answer JSON
- 5. Return the Answer
Files
SKILL.mdcore7,544 bytes
Read SKILL.md
# Northstar Payer Operations Skill
Complete structured prior authorization, appeal, payment integrity, and finance-margin tasks against the Northstar Health Plan shared payer operations environment. For every task, follow the workflow below; return JSON only (no markdown, prose, or commentary outside the JSON object).
---
## 1. Read the Inputs
Read all three task inputs before touching the environment:
| Input | Location | Purpose |
|-------|----------|---------|
| Prompt | `prompt.txt` | Business problem, target ID, role, reporting date, deliverables |
| Task context | `payloads/task_context.json` | Structured metadata: case/claim/queue IDs, dates, domain, review scope |
| Answer template | `payloads/answer_template.json` | Required JSON shape — top-level keys, field types, enums, ordering rules, optional fields flag |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,369 tokensWhen the skill is activated
Reference documents
1,836 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to Use
- Environment Access
- Data Model
- Exploration Workflow
- Answer Construction Rules
- Basis Audit Construction
- Task-Type Specific Guidance
- Common Pitfalls
Files
SKILL.mdcore15,549 bytesSKILL.mdcore15,549 bytesSKILL.mdcore15,549 bytesschema_reference.mddoc6,916 bytes
Read SKILL.md
# Northstar Payer Operations — Structured Determination Skill
## When to Use
Use this skill when tasked with producing a structured JSON determination, summary, or correction packet for a Northstar Health Plan business operation — including utilization management (UM) nurse review, pharmacy appeals intake, payment integrity claim repricing, peer-to-peer (P2P) authorization closure, or UM-finance margin queue analysis.
## Environment Access
All data lives in a shared read-only payer operations environment. Two access modes are available:
### REST Endpoints (no auth header needed)
| Endpoint | Use |
|---|---|
| `GET /api/tables` | Discover the full database schema (table names, columns, types) |
| `GET /api/cases` | List all cases |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,348 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment Setup
- Data Access Strategy
- Business Domain Patterns
- Task-Type Quick Reference
- Numeric Precision
- Answer Construction Checklist
Files
SKILL.mdcore10,595 bytesSKILL.mdcore10,595 bytes
Read SKILL.md
# Northstar Payer Operations Skill
## Overview
Use this skill when solving Northstar Health Plan payer operations tasks — prior authorizations, appeals, payment integrity corrections, peer-to-peer summaries, and margin queue analysis. The environment is a shared read-only payer operations portal with REST endpoints and a SQL interface.
## Environment Setup
The task will provide a base URL and a bearer token. Use these to interact with the environment:
[Credential-related source omitted]
- The SQL endpoint is `POST /sql/query` with JSON body `{"sql": "<statement>"}` — note the key is `sql`, not `query`
- When neither REST nor SQL is specified, prefer the business REST endpoints first; they return pre-joined, comprehensive results
## Data Access Strategy
### Step 1: Use business REST endpoints first
Start with `GET /api/cases/{target_id}` — this returns a rich, pre-joined payload containing the case record, member details, provider info, plan context, request lines, criteria results (with fact traceability), documents (with facts), authorizations, claims with claim lines, appeals, drug trials, P2P events, and assistance screens all in a single call.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,739 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Task Input Structure
- Environment Access Pattern
- Data Model Reference
- General Workflow
- Domain-Specific Patterns
- Output Rules
Files
SKILL.mdcore12,879 bytesSKILL.mdcore12,879 bytes
Read SKILL.md
# Northstar Payer Operations Skill
## Overview
This skill covers Northstar Health Plan payer operations tasks: utilization management determinations, pharmacy coverage appeals, payment integrity corrections, peer-to-peer summaries, and financial queue analysis. Every task uses a shared read-only operations environment and requires a structured JSON answer conforming to a provided template.
## Task Input Structure
Each task provides three inputs:
- **prompt.txt** — Natural-language description of the work item, requester role, reporting date, and what to return.
- **task_context.json** — Machine-readable context: target business IDs, environment access details, and role-specific memos.
- **answer_template.json** — The exact JSON shape to return, including required fields, enum choices, ordering rules, and numeric precision.
## Environment Access Pattern
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,753 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Environment
- Input Dossier
- Workflow
- Data-Quality Rules
- Common Patterns by Work Type
- Error Recovery
- Output Checklist
Files
SKILL.mdcore16,221 bytes
Read SKILL.md
# Northstar Payer Operations — Structured Determination Skill
## Purpose
Produce a structured JSON determination for a Northstar Health Plan payer-operations task by querying the shared read-only environment. Every task follows the same flow: read the input dossier, navigate the environment to collect evidence, apply the correct source-precedence rule for the work type, and return one JSON object matching the supplied answer template.
## When to Use
Invoke this skill when the task:
- Mentions Northstar Health Plan, the shared payer operations environment, or `<TASK_ENV_BASE_URL>`.
- Involves utilization management (UM), prior authorization, appeals, payment integrity / claim repricing, peer-to-peer (P2P) closure, or finance-queue margin analysis.
- Supplies a `prompt.txt`, a `task_context.json` payload, and an `answer_template.json` schema.
- Requires an environment query via REST endpoints or `POST /sql/query`, and a structured JSON-only response.
## Environmentnorthstar-payer-ops
When your agent uses it
Northstar Health Plan payer-operations task execution — prior authorization, appeals, payment integrity, peer-to-peer, and finance margin analysis against the shared payer environment. Produces structured JSON determinations matching an answer template.
What it loads into context
Frontmatter
54 tokensAlways available to the agent
SKILL.md
2,636 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core Operating Rules
Files
SKILL.mdcore12,425 bytesSKILL.mdcore12,425 bytesSKILL.mdcore12,425 bytesSKILL.mdcore12,425 bytes
Read SKILL.md
# Northstar Payer Operations Agent
You are a Northstar Health Plan payer-operations agent. Your job is to receive a task — a prompt, a task context payload, and an answer template — and return exactly one JSON object that conforms to the answer template. You do NOT return markdown, prose, or commentary outside that JSON.
## Core Operating Rules
### 1. Environment Access
All data lives in a shared payer-operations environment. The task prompt supplies a base URL (`<TASK_ENV_BASE_URL>`). Use **only** that environment. Never open, inspect, or read local files — no source files, generated data files, SQLite files, database dumps, manifests, or setup scripts. The environment IS the source of truth.
**SQL access**: `POST /sql/query` to the environment base URL. The authorization header and bearer token are provided in the task materials (context or prompt). Do not hardcode credentials — extract them from the task input.
**Business REST endpoints** — use these when available. The environment typically exposes:
northstar-payer-review
When your agent uses it
Process Northstar Health Plan payer-operations review tasks (prior auth, appeals, payment integrity, P2P, margin queues) against the shared Northstar environment API, producing structured JSON determinations with full basis audit trails.
What it loads into context
Frontmatter
55 tokensAlways available to the agent
SKILL.md
2,302 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
Files
SKILL.mdcore10,989 bytesSKILL.mdcore10,989 bytesSKILL.mdcore10,989 bytesSKILL.mdcore10,989 bytes
Read SKILL.md
# Northstar Payer Operations Review
## Purpose
This skill handles structured review tasks across the Northstar Health Plan payer-operations domain. It covers five review types: utilization management (UM) determinations, pharmacy coverage appeals, payment integrity claim repricing, peer-to-peer (P2P) summaries, and therapy margin queue analysis. The skill produces a single JSON object conforming to the supplied answer template, including a full business basis audit trail.
## When to Use
Invoke this skill when a task requires:
- Reviewing a Northstar case, appeal, claim, P2P event, or margin queue against the shared payer-operations environment
- Producing a structured determination, intake disposition, correction packet, P2P summary, or margin queue report
- The task supplies `task_context.json` (target IDs, requester role, dates, domain) and `answer_template.json` (required output shape)
## Operating RulesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,198 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- General workflow
- Task-type specific guidance
- Distractors and exclusions
- Output rules
Files
SKILL.mdcore15,399 bytesSKILL.mdcore15,399 bytes
Read SKILL.md
# EHR Quality Governance Packet
Use a read-only EHR quality-governance REST API to build normalized JSON packets for clinical-quality workflows: duplicate-chart merge readiness, referral coordination, care transitions, duplicate review with service-request validation, and batch referral audits.
## Environment
The API lives at a base URL provided in the task prompt via the `<TASK_ENV_BASE_URL>` placeholder (or in an `environment_access.md` file if one is present). No authentication is required. The following read-only GET endpoints are available:
| Endpoint | Returns |
|---|---|
| `/api/patients` | List of patients; supports `?name=` and `?mrn=` query params |
| `/api/patients/{patient_id}` | Patient demographics (name, dob, sex, address, phone, insurance, primary_care_provider_id, duplicate_status, etc.) |
| `/api/patients/{patient_id}/conditions` | Active and inactive conditions with ICD-10 codes, descriptions, normalized_keys, status, onset dates |
| `/api/patients/{patient_id}/medications` | Active and inactive medications with names, doses, routes, frequencies, normalized_keys, status |
| `/api/patients/{patient_id}/allergies` | Active and inactive allergies with allergen, reaction, severity, normalized_keys, status |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,743 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Workflow
- API Reference
- Key Conventions
Files
SKILL.mdcore12,350 bytes
Read SKILL.md
# EHR Quality-Governance Skill
## Purpose
Complete EHR quality-governance tasks — duplicate-chart merge readiness packets, referral coordination, care transition packets, duplicate review with ServiceRequest validation, and batch referral audits — by querying a read-only EHR REST API, cross-referencing clinical records, validating codes, and producing normalized JSON output.
## When to Use
Invoke this skill when the task involves:
- A read-only EHR REST API with endpoints for patients, conditions, medications, allergies, encounters, documents, providers, ICD-10 lookup, referrals, duplicates, audit logs, immunizations, disclosures, service requests, and service codes
- Producing a normalized JSON object that conforms to a provided answer template
- Clinical data reconciliation across multiple record sources
- ICD-10 code validation against a directory
- Quality-governance decisions (merge/disposition, referral readiness, audit action plans)
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,034 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Quick-Start Pattern
- API Reference
- Normalization Rules
- Workflow-Specific Patterns
- Error Handling and Edge Cases
- Delivery Checklist
Files
SKILL.mdcore18,597 bytesSKILL.mdcore18,597 bytes
Read SKILL.md
# EHR Quality-Governance API — Reusable Skill
## Overview
This skill teaches how to interact with a shared, read-only EHR (Electronic Health Record) quality-governance API to produce normalized JSON packets for clinical governance workflows. The workflows span duplicate-chart merge readiness, referral coordination, care transitions, duplicate review with service-request quality signals, and batch referral audits.
The API is a RESTful JSON service with ~24 GET endpoints covering patients, clinical lists, encounters, documents, audit logs, duplicate candidates, referrals, ICD-10 codes, providers, and service codes. Every task follows the same pattern: read the prompt, read the answer template, fetch evidence from the API, normalize the data, and return exactly the JSON shape the template demands.
## Quick-Start Pattern
For any new EHR governance task, follow this sequence:
1. **Read the three inputs** — the user's prompt (in `prompt.txt` or inline), the answer template (`answer_template.json`), and any supporting payload files. The template defines the exact output shape and field-level constraints (enums, set semantics, sort orders).
2. **Identify the domain objects** — extract every patient ID, referral ID, candidate ID, service request ID, provider ID, batch ID, and ICD-10 code mentioned in the prompt and payload. These are your API lookup keys.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,049 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Preconditions
- Entry Protocol
- Common Patterns
- Guardrails
Files
SKILL.mdcore14,551 bytesSKILL.mdcore14,551 bytesSKILL.mdcore14,551 bytesSKILL.mdcore14,551 bytesSKILL.mdcore14,551 bytes
Read SKILL.md
# EHR Quality-Governance Skill
## Purpose
Solve structured EHR quality-governance tasks — duplicate-chart merge readiness, referral coordination, care-transition packets, service-request quality review, and referral-batch audits — by reading from a read-only EHR API and producing normalized JSON outputs that conform to an answer template.
## Preconditions
- A base URL (`TASK_ENV_BASE_URL`) for the EHR API is provided in the task.
- Every task includes an answer template (`input/payloads/answer_template.json`) that defines the exact JSON shape, required keys, enum values, and ordering rules.
- Some tasks include supporting payload files (e.g. `merge_packet_request.json`) that name the case objects to work with.
## Entry Protocol
### Step 1 — Read all inputsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,575 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Core Principles
Files
SKILL.mdcore7,777 bytesSKILL.mdcore7,777 bytes
Read SKILL.md
# EHR Quality-Governance Skill
## Purpose
Prepare normalized JSON packets for EHR quality-governance tasks: duplicate-chart merge readiness, referral coordination, care-transition handoffs, duplicate-review with ServiceRequest validation, and referral-batch audits.
## Core Principles
### 1. API-First Data Gathering
Query the EHR environment exhaustively before constructing any answer. Every answer field must be grounded in data returned by the API. Do not guess or interpolate values.
**Standard data-gathering sequence:**
- Read the task prompt to identify the primary entities (patient IDs, referral IDs, duplicate candidate IDs, batch IDs).
- Query the primary entity endpoints first (patient detail, referral detail, duplicate candidate detail).
- Then query all related clinical endpoints (conditions, medications, allergies, encounters, immunizations, documents, service-requests, disclosures).
- Query reference directories (providers, ICD-10 codes, service codes) for every code or provider ID referenced.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,127 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Core workflow
Files
SKILL.mdcore5,815 bytesSKILL.mdcore5,815 bytes
Read SKILL.md
# EHR Quality-Governance Packet Preparation
You are an EHR quality-governance analyst. Given a task prompt and an answer template, query the EHR API environment to gather every relevant piece of evidence, then produce a single normalized JSON object conforming exactly to the answer template.
## Core workflow
### 1. Map the task to endpoints
Read the prompt and the answer template together. Identify every entity mentioned — patient IDs, duplicate candidate IDs, referral IDs, service request IDs, provider IDs, diagnosis codes — and query ALL endpoint families that could return data about them. The available endpoints are documented in the environment; list endpoints give you the full catalogue when you need to discover records by attribute rather than by ID.
**Rule: when a task names a batch, scan the full list endpoint for that batch to discover all member records.** Never assume a batch has a fixed size — count from the API response.
### 2. Gather evidence exhaustively
For every patient involved, fetch: demographics, conditions, medications, allergies, encounters, documents, immunizations, disclosures, and service requests. For every code, validate against the ICD-10 directory. For every provider reference, resolve against the provider directory. For every referral, inspect its detail record. For every duplicate candidate, fetch both the candidate and both patients' full charts.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,978 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
Files
SKILL.mdcore9,182 bytesSKILL.mdcore9,182 bytes
Read SKILL.md
# EHR Quality Governance Skill
## Purpose
Produce a normalized, evidence-based JSON packet from a read-only EHR FHIR-style REST API. The skill covers duplicate-chart merge readiness, referral coordination, care-transition handoff, ServiceRequest quality validation, and batch audit tasks.
## When to Use
Invoke this skill when the task involves:
- An EHR quality-governance or referral-coordination workflow
- A `prompt.txt` describing case objects (patients, duplicate candidates, referrals, ServiceRequests, audit batches)
- An `input/payloads/answer_template.json` specifying the required output JSON shape
- An `environment_access.md` listing the available API base URL and endpoints
## Operating RulesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,698 tokensWhen the skill is activated
Reference documents
2,866 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to apply this skill
- Environment setup
- Core operating rules
- Supporting files
Files
SKILL.mdcore12,707 bytesSKILL.mdcore12,707 bytesapi_reference.mddoc4,718 bytesentity_model.mddoc7,247 bytes
Read SKILL.md
# EHR Quality-Governance Skill
Reusable entry instructions for generating normalized EHR quality-governance packets, referral coordination summaries, care transition packets, duplicate-review validations, and batch audits against a read-only FHIR-aligned REST API. These rules are distilled from five representative task archetypes (merge readiness, referral coordination, care transition, duplicate+ServiceRequest review, and batch audit) and are designed to transfer to any task expressed in the same prompt/payload idiom.
## When to apply this skill
Activate this skill when a prompt:
- References an EHR quality-governance, referral, duplicate-chart merge, care-transition, or audit queue task.
- Points at a read-only REST API reachable at `<TASK_ENV_BASE_URL>` (or a concrete `base_url` provided via `environment_access.md`).
- Asks for normalized JSON output conforming to an `answer_template.json` payload.
- Mentions clinical entities: patients, conditions, medications, allergies, encounters, documents, immunizations, disclosures, providers, referrals, duplicates, ServiceRequests, ICD-10 codes, audit logs, or service codes.
## Environment setup
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,131 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Pre-flight — read these files first
- API query strategy
- Evidence reconciliation rules
- Output construction rules
- Quality checklist
Files
SKILL.mdcore9,616 bytes
Read SKILL.md
# EHR Quality-Governance Skill
## When to use
Invoke this skill when the task involves preparing a normalized JSON packet from an EHR (Electronic Health Record) quality-governance API. The task prompt will reference `<TASK_ENV_BASE_URL>`, an `input/payloads/answer_template.json`, and one or more case objects (patient IDs, duplicate candidate IDs, referral IDs, provider IDs, batch IDs, or service request IDs).
## Pre-flight — read these files first
1. **`environment_access.md`** — Resolves `<TASK_ENV_BASE_URL>` to the actual `base_url`. Lists every `allowed_endpoints` the API exposes. Treat this as the authoritative endpoint catalog; do not call endpoints not listed here.
2. **`input/payloads/answer_template.json`** — The exact JSON shape to return. It defines required top-level keys, field types, enum allowed values, set semantics, and ordering rules. Every field you emit must conform to the constraints in this template.
3. **The task prompt** (`prompt.txt` or equivalent) — Identifies the task type and provides the specific case-object identifiers. Read it to understand which endpoints are relevant and what domain the task falls into.
4. **Any supplementary payloads** in `input/payloads/` beyond `answer_template.json` — Some tasks include additional request objects (e.g., `merge_packet_request.json`) that carry structured parameters. Read all payload files present; do not assume only the answer template exists.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,209 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Input Files
- Clinic Runtime API
- Data Retrieval Workflow
- Response Construction Rules
- General Constraints
Files
SKILL.mdcore9,873 bytes
Read SKILL.md
# Synthetic Clinic Protocol Decision Support Skill
## Purpose
This skill produces a structured, protocol-bound clinical decision-support JSON response for a synthetic clinic case. It retrieves clinical data from a read-only clinic runtime REST API, cross-references that data with applicable protocol materials, and returns a single JSON object that conforms exactly to the supplied answer template schema.
## Input Files
Before making any API calls, read these three input files in the task directory:
1. **`input/prompt.txt`** — Identifies the target case ID, the clinical domain (respiratory, head injury, electrolyte replacement, care management, lab-window retrieval, etc.), and any domain-specific requirements. It also names the answer template to use.
2. **`input/payloads/answer_template.json`** — The output schema. It defines:
- Required top-level keys
- Allowed enum values for every controlled fieldWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,943 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use this skill
- Core workflow
- Anti-patterns
Files
SKILL.mdcore8,752 bytes
Read SKILL.md
# Clinic Decision-Support Skill
Use the synthetic clinic runtime environment to produce structured, protocol-bound
clinical decision-support answers from patient data exposed through a REST API.
## When to use this skill
The task will provide:
- A **prompt** naming a target case identifier and describing the clinical question.
- An **answer template** (`input/payloads/answer_template.json`) defining the exact
JSON output shape: required top-level keys, field types, allowed enum values,
numeric precision rules, and safety-check requirements.
- **Environment access details** (`environment_access.md` or equivalent) listing
the base URL, credentials, and allowed API endpoints for the run.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,807 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Read the environment access file
- 2. Read the task prompt
- 3. Read the answer template
- 4. Query the clinic API
- 5. Apply clinical reasoning
- 6. Return the answer
- Common clinical patterns
- Authentication
Files
SKILL.mdcore8,075 bytesSKILL.mdcore8,075 bytesSKILL.mdcore8,075 bytes
Read SKILL.md
# Synthetic Clinic Decision-Support Skill
You are an agent that produces structured clinical decision-support responses for synthetic clinic cases. You work against a read-only FHIR-like REST API that models a clinic runtime environment.
## 1. Read the environment access file
Locate `environment_access.md` in the working directory. It contains:
- `base_url` — the root URL of the clinic runtime (e.g. `http://localhost:9016/`).
[Credential-related source omitted]
- `allowed_endpoints` — the permitted HTTP methods and paths (e.g. `GET /api/patients`). Do not call endpoints outside this list.
## 2. Read the task prompt
Read `input/prompt.txt` (or whatever path the prompt lives at for this run). The prompt states:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,350 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Orient to the task
- 2. Retrieve and apply the protocol
- 3. Filter and match observations rigorously
- 4. Respect window boundaries
- 5. Build the answer
- 6. Cross-validate before finalising
Files
SKILL.mdcore6,306 bytesSKILL.mdcore6,306 bytesSKILL.mdcore6,306 bytes
Read SKILL.md
# Clinic Decision-Support Skill
Apply this skill when solving structured clinical decision-support tasks against a synthetic
clinic runtime environment. The tasks require retrieving patient case data, applying protocol
rules, and returning JSON answers that conform to a supplied template.
## 1. Orient to the task
Read the prompt first. It names the target case identifier, describes the clinical domain, and
points to an answer template. Before fetching anything, read the answer template to understand
every required key, its type, allowed enum values, and any ordering or nullability rules. The
template is the single source of truth for the output shape.
Then fetch the target case. The case endpoint returns a composite payload that bundles the
patient record, encounter findings, observations, medications, allergies, imaging, problems,When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,835 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Workflow
Files
SKILL.mdcore8,297 bytes
Read SKILL.md
# Clinic Protocol Decision Support Skill
## Purpose
Solve structured clinical decision-support tasks that require retrieving synthetic patient data from a clinic runtime API, applying protocol rules to the retrieved facts, and returning a single JSON object that conforms to a provided answer-template schema.
## When to Use
The task prompt will direct you to a synthetic clinic runtime environment, name a target case, reference an answer-template file, and ask for a protocol-bound clinical assessment or routing decision. Use this skill whenever you see that pattern.
## Workflow
### Phase 1 — Gather All Available Data
Before reasoning about the case, retrieve every resource the clinic environment exposes. Cast a wide net first; narrow to the target case second.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,675 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Core Workflow
- Common Pitfalls
Files
SKILL.mdcore7,897 bytesSKILL.mdcore7,897 bytes
Read SKILL.md
# Clinic Decision-Support Skill
## Purpose
Generate protocol-bound clinical decision-support JSON responses for synthetic clinic cases. Each task presents a case identifier, an answer template, and access to a clinic runtime environment with patient records, observations, protocols, and related clinical data.
## Core Workflow
Follow these steps in order for every task:
### 1. Orient to the Task
Read the task prompt to identify:
- The target case identifier (e.g., `CASE-XXX-NNN`)
- The answer template file at `input/payloads/answer_template.json`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,944 tokensWhen the skill is activated
Reference documents
2,138 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Operating rules
- Reusable patterns
- Reference files
Files
SKILL.mdcore9,229 bytesSKILL.mdcore9,229 bytesreferences/api-surface.mddoc4,620 bytesreferences/template-format.mddoc4,669 bytes
Read SKILL.md
# Clinic Decision Support Skill
Produce protocol-bound clinical decision-support responses by querying a synthetic clinic runtime API and returning a single structured JSON object conforming to a supplied answer template.
## When to use
Invoke this skill when the task involves:
- A clinic case identifier (e.g., `CASE-*-NNN`) in the prompt
- A structured answer template (`answer_template.json`) in the task payloads
- A clinic runtime environment described in a separate environment-access file
- A requirement to return exactly one JSON object with controlled vocabulary
## Operating rules
### Phase 1: Load environment configurationWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,280 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Step-by-Step Procedure
- Common Clinical Decision Patterns
Files
SKILL.mdcore10,371 bytes
Read SKILL.md
# Synthetic Clinic Clinical Decision-Support Skill
## Purpose
This skill guides an agent through completing a structured clinical decision-support task against a synthetic clinic FHIR-like REST API. The agent reads a task prompt, retrieves structured clinic data from the runtime environment, and returns a single JSON object conforming to the provided answer template.
## When to Use
Invoke this skill when:
- A task directory under `train_tasks/train_NNN/input/` contains `prompt.txt` and `payloads/answer_template.json`.
- A file named `environment_access.md` is present at the workspace root providing the runtime base URL, credentials, and allowed endpoints.
## Step-by-Step Procedure
### Step 1 — Discover the InputsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,213 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
- Task-Type Specific Guidance
- Integration Pattern
Files
SKILL.mdcore10,649 bytesSKILL.mdcore10,649 bytes
Read SKILL.md
# Synthetic Clinic Protocol Decision-Support Skill
## Purpose
Execute clinical protocol decision-support tasks against a synthetic clinic runtime environment. This skill covers reading a structured task prompt and answer template, querying the clinic API for case and patient data, applying protocol rules to determine assessments and recommendations, and returning a strict JSON output with no extraneous text.
## When to Use
Invoke this skill when the task involves:
- A synthetic clinic case identifier (e.g., `CASE-RESP-102`, `CASE-HEAD-207`, `CASE-K-303`, `CASE-CM-411`, `CASE-LAB-518`)
- A runtime environment reference (`<TASK_ENV_BASE_URL>`)
- A structured JSON answer template (`input/payloads/answer_template.json`)
- Clinical decision-support domains: respiratory protocol, head injury, potassium replacement, care management routing, or lab observation retrieval
## Operating RulesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,966 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Environment Access
- Available Endpoints
- General Workflow
- Common Enum Families (Conceptual)
- SQL Query Patterns
- Anti-Patterns (Do Not Do)
- Quick-Start Checklist
Files
SKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytesSKILL.mdcore12,708 bytes
Read SKILL.md
# Investigation Review Hub — Structured Assessment Skill
## Purpose
Produce a single, conformant JSON assessment for a legal investigation matter by
querying the shared Investigation Review Hub REST and SQL endpoints. The hub is
the **sole source of record** for matter metadata, subpoena/request categories,
production status, custodian sources, review documents, privilege-log data, QC
findings, retention events, and remediation actions.
## Environment Access
Read the task-provided `environment_access.md` (or equivalent) for:
- **`base_url`** — the root URL of the Investigation Review Hub (e.g.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,147 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Discover the environment
- 2. Understand the task
- 3. Master the answer template
- 4. Query the hub
- 5. Assemble the deliverable
- 6. Return format
Files
SKILL.mdcore8,894 bytesSKILL.mdcore8,894 bytesSKILL.mdcore8,894 bytesSKILL.mdcore8,894 bytesSKILL.mdcore8,894 bytes
Read SKILL.md
# Investigation Review Hub — Structured JSON Deliverable
Use the shared **Investigation Review Hub** as the exclusive source of record to
produce a structured JSON deliverable for an eDiscovery / legal-investigation
matter (gap analysis, retention review, remediation dashboard,
production-readiness review, or a similar workstream).
## 1. Discover the environment
Read `environment_access.md` for the live base URL and API credentials.
The file supplies:
- `base_url` — the root of the Investigation Review Hub (e.g. `http://task-env:9017/`)
- `credentials.sql_endpoint_header` — the HTTP header name used to authenticate
read-only SQL queries (e.g. `X-API-Key`)When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,590 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Input Anatomy
- 2. Hub API Access
- 3. Workflow
- 4. Common Evidence Patterns
- 5. Action Planning Conventions
- 6. SQL Query Tips
Files
SKILL.mdcore10,928 bytesSKILL.mdcore10,928 bytesSKILL.mdcore10,928 bytesSKILL.mdcore10,928 bytes
Read SKILL.md
# Investigation Review Hub — Structured Legal-Review JSON Deliverable
Use the **Investigation Review Hub** REST+SQL API to produce structured JSON
deliverables for eDiscovery, privilege review, production readiness, retention
gap analysis, and remediation dashboards. Every deliverable conforms to a
task-supplied `answer_template.json` schema; evidence comes exclusively from
Hub endpoints, never from local environment files, database files, or manifests.
---
## 1. Input Anatomy
Each task directory contains:
| Path | Role |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,409 tokensWhen the skill is activated
Reference documents
678 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to Use
- Hub Access
- Schema Reference
- Separating Signal from Noise
- Building the Answer
- Common Pitfalls
Files
SKILL.mdcore10,899 bytesSKILL.mdcore10,899 bytesSKILL.mdcore10,899 bytesSKILL.mdcore10,899 bytesschema_reference.mddoc2,966 bytes
Read SKILL.md
# Investigation Review Hub — Structured Gap Analysis Skill
## When to Use
Invoke this skill when the task involves an Investigation Review Hub matter and asks for a structured JSON deliverable: a gap analysis, remediation dashboard, retention review, privilege review, or production-readiness assessment. The hub holds multiple matters; always confirm the `matter_id` from the task payload before querying.
## Hub Access
Use the task-provided base URL (typically `http://task-env:9017/`). All read operations go through two channels:
**Dedicated GET endpoints** (no auth header required for read-only endpoints):
- `GET /api/matters` — list all matters
- `GET /api/subpoena-categories` — request categories per matter
- `GET /api/productions` — production batch stats
- `GET /api/custodian-sources` — custodian devices, archives, collection statusWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,870 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Data Sources
- Core Workflow
- Common Pitfalls
Files
SKILL.mdcore8,265 bytesSKILL.mdcore8,265 bytesSKILL.mdcore8,265 bytesSKILL.mdcore8,265 bytesSKILL.mdcore8,265 bytesSKILL.mdcore8,265 bytes
Read SKILL.md
# Investigation Review Hub — Structured Analysis Skill
## Purpose
Produce structured JSON deliverables for legal-matter investigations (gap analysis, retention review, remediation dashboard, production-readiness review) using a read-only Investigation Review Hub as the sole source of record.
## Data Sources
The Investigation Review Hub exposes these tables via a read-only SQL endpoint:
| Table | Key columns | Purpose |
|---|---|---|
| `matters` | matter_id, name, agency, hold_date, investigation_type, status | Matter metadata and hold date |
| `subpoena_categories` | matter_id, category_code, title, topic_tags | Request categories with scope dates |
| `production_stats` | matter_id, batch_id, category_code, produced_count, withheld_count, responsive_count, status | Production batch statistics per category |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,679 tokensWhen the skill is activated
Reference documents
1,605 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Phase 1 — Inventory the input directory
- Phase 2 — Discover the environment base URL
- Phase 3 — Retrieve matter data from the hub
- Phase 4 — Map hub values to template enums
- Phase 5 — Construct the answer
- Phase 6 — Validate before output
- Common pitfalls
- Supporting files
Files
SCHEMA_REFERENCE.mddoc6,304 bytesSKILL.mdcore11,139 bytesSKILL.mdcore11,139 bytesSKILL.mdcore11,139 bytes
Read SKILL.md
# Investigation Review Hub — Structured JSON Deliverable Skill
## When to use
Use this skill when a task prompt asks you to produce a structured JSON deliverable from an **Investigation Review Hub** — a read-only REST API serving matter metadata, subpoena categories, production stats, custodian sources, review documents, privilege logs, QC findings, retention events, and remediation actions. The prompt will reference `<TASK_ENV_BASE_URL>` and provide an `answer_template.json` schema. Return **only** the JSON object; no surrounding prose.
## Phase 1 — Inventory the input directory
Every task instance ships with these files inside `input/`:
| File | What it carries |
|---|---|
| `prompt.txt` | Business narrative, matter name, deliverable type, `<TASK_ENV_BASE_URL>` placeholder |
| `payloads/answer_template.json` | Required JSON schema: top-level keys, per-item required keys, enum choices, ordering rules, numeric precision |
| `payloads/request_context.json` *or* `review_scope.json` *or* `matter_context.json` | The `matter_id` string, client-facing category labels, any scope notes |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,983 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment Resolution
- 2. API Surface
- 3. Source Discipline (Non-Negotiable)
- 4. Task Input Anatomy
- 5. Output Discipline
- 6. Common Answer Schema Patterns
- 7. Workflow: Step-by-Step Investigation Sequence
- 8. Domain Vocabulary Reference
- 9. Error Prevention Checklist
Files
SKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytesSKILL.mdcore12,857 bytes
Read SKILL.md
# Investigation Review Hub — Agent Operating Rules
These are the reusable operating rules for working with the Investigation Review
Hub, distilled from gap-analysis, retention-review, remediation-dashboard, and
production-readiness tasks. Follow these rules whenever a task directs you to
the Hub.
## 1. Environment Resolution
- The task prompt and payloads reference the Hub with the placeholder
`<TASK_ENV_BASE_URL>`. Resolve this by reading the task-local file
`environment_access.md` (or, if absent, the `base_url` / `environment` block
in the task's context payload) to obtain the actual base URL and credentials.
- The standard read-only API key header is `X-API-Key`; the key value is
`review-key-017` unless the task provides different credentials.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,464 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
Files
SKILL.mdcore11,072 bytesSKILL.mdcore11,072 bytesSKILL.mdcore11,072 bytesSKILL.mdcore11,072 bytesSKILL.mdcore11,072 bytesSKILL.mdcore11,072 bytes
Read SKILL.md
# Investigation Review Hub — Gap Analysis & Remediation Dashboard
## Purpose
Generate a structured JSON deliverable for a legal investigation review by querying a shared Investigation Review Hub REST API, mapping its evidence (matters, subpoena categories, productions, custodian sources, documents, privilege-log entries, QC findings, retention events, and remediation actions) into a formal answer that conforms to a task-supplied JSON answer template.
## When to Use
The task prompt asks you to use an Investigation Review Hub at a configurable base URL and requests a structured JSON output — typically a gap analysis, retention/preservation review, production-readiness assessment, remediation dashboard, or priority action plan tied to subpoena/request categories.
---
## Operating Rules
### 1. Resolve the EnvironmentWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,754 tokensWhen the skill is activated
Reference documents
4,049 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
- Task Archetypes
- Workflow
- Common Enum Reference
- Supporting Files
Files
SKILL.mdcore16,532 bytesSKILL.mdcore16,532 bytesSKILL.mdcore16,532 bytesSKILL.mdcore16,532 bytesSKILL.mdcore16,532 bytesreferences/api-endpoints.mddoc8,810 bytesreferences/enum-usage.mddoc9,293 bytes
Read SKILL.md
# Investigation Review Hub — Structured Review Skill
## Purpose
Generate structured JSON review deliverables for legal e-discovery matters by querying the Investigation Review Hub API and conforming to task-provided answer templates. Covers gap analyses, retention reviews, remediation dashboards, and production-readiness assessments.
## When to Use
Invoke this skill when the task involves:
- An Investigation Review Hub at a resolvable base URL
- A `prompt.txt` describing the review deliverable
- An `input/payloads/answer_template.json` defining the output schema
- Legal e-discovery review types: gap analysis, retention/hold review, remediation dashboard, or production readiness
## Operating RulesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,040 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- When to Use
- Workflow
- Portal Query Patterns
- Reconciliation Rules Reference
- Common Task Patterns
- Output Validation Checklist
Files
SKILL.mdcore18,080 bytesSKILL.mdcore18,080 bytesSKILL.mdcore18,080 bytesSKILL.mdcore18,080 bytes
Read SKILL.md
# Court Clerk Post-Hearing Reconciliation and Financial Packet Skill
## Overview
This skill prepares structured post-hearing or post-disposition reconciliation packets for
court clerk entry. It cross-references local case materials (hearing notes, clerk memos,
finance queue extracts, petition summaries, sentencing intake sheets, worksheet CSVs, and
local form excerpts) against the **Court Operations Portal** (`<TASK_ENV_BASE_URL>`) to
produce a single JSON answer conforming to an attached `answer_template.json`.
The skill handles criminal sentencing closeouts, traffic-violation payment-plan packets,
post-sentencing field packets (including CC-1375 probation referrals and CC-1379 license
suspension / installment orders), and multi-case criminal disposition registers.
## When to UseWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,533 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Workflow
- Common Resolution Patterns (Reference)
- Portal Interaction Notes
Files
SKILL.mdcore12,018 bytesSKILL.mdcore12,018 bytesSKILL.mdcore12,018 bytesSKILL.mdcore12,018 bytes
Read SKILL.md
# Court Clerk Closeout & Field Packet Skill
## Purpose
Prepare structured clerk-ready JSON outputs for court disposition closeouts, post-sentencing field packets, and financial reconciliation batches. This skill covers criminal, traffic, and civil docket processing where local case materials must be cross-referenced against a centralized Court Operations Portal to resolve conflicts and produce a single authoritative answer.
## When to Use
Invoke this skill when the task involves:
- Closing a criminal or traffic docket batch from hearing notes and finance queue extracts
- Preparing post-sentencing field packets with probation referrals, license suspension orders, and payment plans
- Reconciling financial entries against current fee schedules from a court operations portal
- Identifying audit conflicts between multiple data sources (finance queue, hearing notes, CMS, worksheets, audit memos)
- Producing structured JSON answers matching a court-provided answer template with enums, ISO dates, and two-decimal currency
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,346 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Phase 1 — Gather Inputs
- Phase 2 — Cross-Reference and Identify Conflicts
- Phase 3 — Build the Answer
- Phase 4 — Validate Before Returning
- Portal API Reference
- Common Patterns by Task Type
- Quick Reference: Do and Don't
Files
SKILL.mdcore14,677 bytesSKILL.mdcore14,677 bytes
Read SKILL.md
# Court Clerk Closeout Packet Skill
You are a deputy court clerk preparing case closeout packets, post-sentencing field packets, or disposition batches. Your job is to cross-reference local case materials with the Court Operations Portal, resolve conflicts, and produce a structured JSON answer that a clerk can enter directly into the case management system.
## Phase 1 — Gather Inputs
### 1a. Read the task prompt
The prompt (typically `input/prompt.txt`) names:
- The court and jurisdiction
- The target cases, citations, or petitions
- Which portal endpoints are relevant
- Which local payloads to use
### 1b. Read all local payloadsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,872 tokensWhen the skill is activated
Reference documents
899 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Entry Conditions
- Core Workflow
- Common Patterns
- API Reference
- Edge Cases and Defensive Checks
- What NOT to Do
Files
SKILL.mdcore13,084 bytesreference.mddoc3,937 bytes
Read SKILL.md
# Court Clerk Operations — Multi-Source Reconciliation
## Purpose
Process court closeout packets, disposition registers, and post-sentencing field packets by cross-referencing local case materials against a Court Operations Portal API. Produce structured JSON outputs that reconcile conflicts across data sources, apply current fee schedules, and flag items that must be held, excluded, or verified before entry.
## Entry Conditions
Invoke this skill when the task involves:
- Closing out a criminal docket, traffic docket, or disposition batch
- Preparing a post-sentencing field packet or financial/supervision packet
- Reconciling a finance queue or clerk worksheet against court records
- Any task that directs you to cross-reference local payloads with a "Court Operations Portal" API
## Core WorkflowWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,112 tokensWhen the skill is activated
Reference documents
1,304 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to Use This Skill
- Core Methodology
- Cross-Domain Patterns
- Workflow Summary
Files
SKILL.mdcore9,680 bytesSKILL.mdcore9,680 bytesSKILL.mdcore9,680 bytesSKILL.mdcore9,680 bytesreferences/fee-rules.mddoc2,241 bytesreferences/payment-plan-math.mddoc1,749 bytesreferences/source-hierarchy.mddoc1,726 bytes
Read SKILL.md
# Court Operations Reconciliation Skill
## When to Use This Skill
Use this skill when asked to prepare a court clerk closeout, disposition register, post-sentencing packet, or financial reconciliation across multiple case matters. The skill covers reconciling court records from multiple sources — case management systems (CMS), hearing notes, financial queues, audit memos, and payment petitions — into a single structured clerk-ready output.
---
## Core Methodology
### 1. Gather All Sources Before Deciding
Read every provided source document and query the portal for every relevant case, charge, docket entry, fee schedule, form, payment policy, and petition. Never base a decision on one source alone. The sources to collect include:
- **Portal/CMS**: case records, charge records, docket entries, fee schedules, payment policies, forms, petitions, citationsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,044 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Core Workflow
- Domain-Specific Patterns
- Output Rules
Files
SKILL.mdcore14,441 bytesSKILL.mdcore14,441 bytesSKILL.mdcore14,441 bytesSKILL.mdcore14,441 bytesSKILL.mdcore14,441 bytes
Read SKILL.md
# Court Closeout Clerk — Structured Reconciliation Skill
## Purpose
You are a deputy clerk preparing court closeout packages. Given a task prompt, local case-file payloads, and access to a Court Operations Portal, produce a single structured JSON answer matching the provided answer template. Your work reconciles conflicts across data sources, applies current fee schedules and payment policies, and flags items that cannot be resolved.
## Core Workflow
Follow these steps in order for every task. Never skip a step or assume data is clean before you verify it.
### Phase 1 — Orient
1. **Read the prompt.** Identify the target court, jurisdiction, hearing/docket date, and the list of target case numbers, citation numbers, or petition IDs. Note which portal endpoints are listed as available.
2. **Read every file in the payloads directory.** There will always be an `answer_template.json` — read it first to understand the required output shape, required top-level keys, enums, ordering rules, currency precision, and date format. Other payloads (hearing notes, audit memos, finance extracts, petition summaries, budget worksheets, form-field excerpts, sentencing intake sheets) contain the ground-truth facts for each matter.
3. **Note the jurisdiction code** for each target matter. This determines which fee schedules, payment policies, and forms apply. You will need it to filter portal results.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,487 tokensWhen the skill is activated
Reference documents
626 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to use
- Workflow
- Operating principles
- Supporting files
Files
SKILL.mdcore11,575 bytesSKILL.mdcore11,575 bytesSKILL.mdcore11,575 bytesportal-api-reference.mddoc2,950 bytes
Read SKILL.md
# Court Clerk Operations — Closeout and Financial Reconciliation
## Purpose
Prepare court closeout, disposition, and financial-reconciliation packets by cross-referencing local case materials against the Court Operations Portal API and producing a structured JSON answer that follows an answer-template schema.
## When to use
The user is acting as a deputy clerk or court officer and needs to:
- Reconcile hearing notes, audit memos, finance queue extracts, or petition summaries against the live case-management system.
- Produce a clerk-ready JSON output matching a provided answer-template schema.
- Identify audit conflicts, stale fee amounts, unsupported charges, and missing identifiers.
- Decide which cases should be **posted** (disposed with signed orders) and which must be **held** (pending, continued, deferred, or unsigned).
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,108 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to use
- Base URL and network access
- Available API endpoints
- Input structure
- Master workflow
- Reconciliation rules
- Payment plan rules
- Form metadata rules
- Placeholder handling
- Exclusions
- Enum discipline
- Formatting rules
- Jurisdiction contexts
- Answer delivery
Files
SKILL.mdcore14,795 bytes
Read SKILL.md
# Court Operations Portal — Clerk Closeout & Packet Preparation
## Purpose
Process court clerk closeout, sentencing packet, disposition register, and post-disposition financial/supervision packet tasks. These tasks combine local payloads (hearing notes, clerk memos, finance extracts, form excerpts, petition summaries) with the Court Operations Portal REST API to produce a structured JSON answer matching a provided answer template.
## When to use
- The task asks you to act as a court clerk (deputy clerk, clerk, or similar court-official role)
- The task references a Court Operations Portal at `<TASK_ENV_BASE_URL>`
- The task provides an `answer_template.json` in an `input/payloads/` directory
- The task supplies local payloads such as hearing notes, clerk audit memos, finance queue extracts, form field excerpts, petition summaries, or sentencing/probation intake notes
- The prompt lists a set of case numbers, citation numbers, or petition IDs to close out or process
## Base URL and network accessWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,115 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Environment Setup
- Input Materials
- Output Rules
- Core Operating Rules
Files
SKILL.mdcore19,117 bytes
Read SKILL.md
# Court Clerk Disposition Closeout Skill
## Purpose
Perform court clerk disposition-closeout tasks for criminal and traffic dockets. Given local case materials and API access to a Court Operations Portal, produce a single JSON answer conforming to the provided answer template schema. Reconcile hearing notes, clerk memos, intake sheets, finance extracts, and portal records into a clerk-ready register entry.
## Environment Setup
### Base URL
All API calls use `http://task-env:9018/` as the base URL. Never use `localhost`, `127.0.0.1`, or any URL from an `env/setup.sh` script — the base URL from the task environment always takes precedence.
### Authentication
No credentials are required. All endpoints are unauthenticated GET requests.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,817 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Step 1 — Read environment configuration
- Step 2 — Read the task prompt
- Step 3 — Read the answer template
- Step 4 — Fetch all relevant data
- Step 5 — Cross-reference records
- Step 6 — Apply business rules
- Step 7 — Assemble the output
- Step 8 — Validate before returning
Files
SKILL.mdcore8,434 bytes
Read SKILL.md
# Licensing Review Agent
You are a licensing review agent that processes regulatory application batches, produces
structured decisions, and builds staff review packages by querying a shared licensing
data environment over HTTP.
## Step 1 — Read environment configuration
Locate and read `environment_access.md` at the top level of the working directory.
Extract:
- `base_url` — the root URL of the licensing data service.
- `credentials` — any required auth headers (token name, value) and which endpoint
they apply to. Apply these to every request that needs them.
- `allowed_endpoints` — the exact list of endpoints you are permitted to call. DoWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,391 tokensWhen the skill is activated
Reference documents
5,045 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Phase 1 — Environment Bootstrapping
- Phase 2 — Task Discovery
- Phase 3 — Data Collection
- Phase 4 — Decision Logic
- Phase 5 — Output Construction
- Phase 6 — Verification (Before Returning)
- Reference Files
Files
SKILL.mdcore10,508 bytesreferences/domain-concepts.mddoc9,767 bytesreferences/task-patterns.mddoc11,704 bytes
Read SKILL.md
# Licensing Examiner Skill
Reusable methodology for evaluating licensing applications, building staff review
packages, and constructing manual-review queues against a shared licensing data
environment. This skill covers contractor eligibility batches, restricted liquor
license transfers, and alcohol renewal screening — any task where you must fetch
records from the licensing API, cross-reference them against business rules, and
produce structured JSON decisions.
---
## Phase 1 — Environment Bootstrapping
Every task begins by locating and reading the environment descriptor.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,003 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment Setup
- 2. Task-Type Recognition
- 3. Data Collection Strategy
- 4. Domain Analysis Frameworks
- 5. Output Formatting Rules
- 6. Step-by-Step Execution Protocol
Files
SKILL.mdcore22,335 bytes
Read SKILL.md
# Licensing Review Skill
Reusable instructions for processing structured licensing-review tasks in contractor, liquor, and alcohol-renewal domains. This skill covers environment setup, task-type recognition, data collection, domain analysis, and output formatting.
---
## 1. Environment Setup
Every task references `<TASK_ENV_BASE_URL>` in the prompt. Resolve this by reading the file `environment_access.md` at the workspace root, which contains:
- **`base_url`** — the actual HTTP origin to use in place of `<TASK_ENV_BASE_URL>`.
- **`credentials`** — a mapping of HTTP method + path pattern to required headers. The most common credential is a header `X-Task-Token` required on `POST /api/sql` calls.
- **`allowed_endpoints`** — the full list of endpoints available in this environment instance. Only endpoints listed here (and in the prompt) should be called.
### Making API callsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,698 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Task Categories
- General Workflow
- Pattern Reference
Files
SKILL.mdcore12,682 bytes
Read SKILL.md
# Licensing Review Skill
## Purpose
Solve structured licensing review tasks by querying a shared licensing environment, cross-referencing entity records against active policies, and producing JSON answers conforming to a supplied answer template.
## Task Categories
There are three recurring task patterns:
### 1. Contractor Batch Eligibility Review
Review a batch of contractor applications against the State Contractors Licensing Board policies. Each application must be evaluated for bond, insurance, experience, endorsement, license history, violations, and inspection compliance.
### 2. Liquor License Staff Package
Prepare a single-application staff review package for a restricted liquor license (transfer, renewal with controls, or new). Evaluate settlements at the target location, active controls, incidents, site evidence, and applicable privilege obligations.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,496 tokensWhen the skill is activated
Reference documents
1,161 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Step 1 — Locate the Environment
- Step 2 — Read the Answer Template
- Step 3 — Fetch All Relevant Data
- Step 4 — Apply Policy Rules
- Step 5 — Liquor License Staff Package Rules
- Step 6 — Alcohol Renewal Queue Rules
- Step 7 — Assemble and Validate
Files
SKILL.mdcore15,061 bytesreference.mddoc4,368 bytes
Read SKILL.md
# Licensing Review Skill
You are a senior licensing examiner. When given a batch of applications, produce a structured JSON decision package by fetching data from the licensing environment, applying the governing policies, and rendering your answer in the exact schema provided.
---
## Step 1 — Locate the Environment
Read `environment_access.md` (or equivalent environment instructions provided with the task) to obtain:
- `base_url` — the task environment root (e.g., `http://task-env:9019/`).
[Credential-related source omitted]
- `allowed_endpoints` — the complete list of available GET and POST endpoints.
If the task prompt mentions `<TASK_ENV_BASE_URL>`, substitute it with the actual `base_url`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,745 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Phase 1 — Understand the Task
- Phase 2 — Fetch Records
- Phase 3 — Classify Each Application
- Phase 4 — Verify the Output
- General Principles
Files
SKILL.mdcore12,278 bytesSKILL.mdcore12,278 bytes
Read SKILL.md
# Licensing Review Skill
You are a licensing examiner supporting state regulatory boards. When given a batch of applications, follow these instructions to produce a structured JSON decision package.
## Phase 1 — Understand the Task
1. **Read the prompt** to identify the domain (contractor, liquor, or alcohol renewal), the target application IDs, any review date, and which endpoints are available.
2. **Read the answer template** (`input/payloads/answer_template.json`) before fetching any data. Note every required key, allowed enum values, ordering rules, and whether lists must be sorted. The template is the contract — your output must conform exactly.
## Phase 2 — Fetch Records
Use the shared licensing environment. The base URL is provided in environment instructions as `<TASK_ENV_BASE_URL>`. The available GET endpoints vary by domain:
- **Contractor tasks**: `/api/policies`, `/api/contractor/applications`, `/api/contractor/bonds`, `/api/contractor/insurance`, `/api/contractor/license-history`, `/api/contractor/violations`, `/api/contractor/correspondence`, `/api/contractor/inspections`
- **Liquor tasks**: `/api/policies`, `/api/liquor/applications`, `/api/liquor/settlements`, `/api/liquor/privileges`, `/api/liquor/incidents`, `/api/liquor/site-evidence`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,080 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Operating Rules
Files
SKILL.mdcore9,381 bytes
Read SKILL.md
# Licensing Review Skill
## Purpose
Conduct structured eligibility and compliance reviews for state licensing board applications — contractor, liquor, and alcohol renewal — by fetching data from a shared licensing API environment, applying business rules, and returning schema-compliant JSON decisions.
## When to Use
Invoke this skill when a prompt assigns a licensing examiner role and references a task environment base URL (`<TASK_ENV_BASE_URL>`), a batch of application or license identifiers, and an answer template schema.
---
## Operating Rules
### 1. Environment Bootstrap
1. Locate `environment_access.md` in the working directory.
2. Extract `base_url`. Substitute every occurrence of `<TASK_ENV_BASE_URL>` in the prompt with this value.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,785 tokensWhen the skill is activated
Reference documents
925 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Operating Rules
Files
SKILL.mdcore12,800 bytesreference/code_mappings.mddoc3,796 bytes
Read SKILL.md
# Licensing Review Skill
## Purpose
Complete structured licensing review tasks by interacting with a task-environment API. This skill covers contractor eligibility batches, liquor license staff packages, and alcohol renewal queues — any task that supplies a prompt with target applications/licensees, a set of API endpoints, and an answer template.
## Operating Rules
### Phase 1 — Environment Setup
1. Locate `environment_access.md` in the working directory. It provides:
- `base_url` — the root URL for all API calls (referenced as `<TASK_ENV_BASE_URL>` in prompts).
- `credentials` — any required headers (e.g., `X-Task-Token`) and which endpoints require them.
- `allowed_endpoints` — the full list of callable endpoints.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,085 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Phase 1: Environment Bootstrap
- Phase 2: Task Intake
- Phase 3: Data Collection
- Phase 4: Domain Decision Rules
- Phase 5: Output Construction
- Phase 6: Self-Check Before Returning
Files
SKILL.mdcore14,116 bytesSKILL.mdcore14,116 bytes
Read SKILL.md
# Licensing Review Skill
You are a structured licensing decision engine. When invoked, you operate in a standardized licensing review environment. Follow these phases in order.
---
## Phase 1: Environment Bootstrap
1. Read `environment_access.md` from the working directory to obtain:
- `base_url` — the root URL of the licensing data service (substitute this wherever you see `<TASK_ENV_BASE_URL>` in prompts).
- `credentials` — any required headers (e.g., `X-Task-Token`) and the endpoints they apply to. Apply these headers only when calling the specified methods and paths.
- `allowed_endpoints` — the full list of available API routes. Prefer GET endpoints for data collection; use POST `/api/sql` only when GET endpoints cannot answer a question and the credentials grant SQL access.
2. If `environment_access.md` is missing or does not declare a `base_url`, stop and report the gap — do not guess.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,761 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment
- General Workflow
- Task-Type Specific Guidance
- Common Pitfalls to Avoid
Files
SKILL.mdcore12,706 bytesSKILL.mdcore12,706 bytesSKILL.mdcore12,706 bytesSKILL.mdcore12,706 bytesSKILL.mdcore12,706 bytes
Read SKILL.md
# M&A Deal Workbench Review Skill
## Overview
This skill provides a reusable workflow for reviewing M&A transaction records using a deal workbench REST API. It covers seller-side and buyer-side deal review, committee escalation, carveout transition review, and deviation matrices across asset purchases, stock purchases, and mergers.
## Environment
The deal workbench is a running REST API. Its base URL is provided as `<TASK_ENV_BASE_URL>` in task prompts (typically `http://task-env:9020/`). A read-only SQL endpoint is available at `POST /api/query` with bearer token `deal-workbench-readonly`. All GET endpoints are unauthenticated.
### Available API Endpoints
**Deal-level records:**
- `GET /api/deals/<deal_id>` — deal metadata, headline purchase price, parties, structure
- `GET /api/deals/<deal_id>/terms` — current draft agreement termsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,147 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Activation
- Core Workflow
Files
SKILL.mdcore9,438 bytesSKILL.mdcore9,438 bytes
Read SKILL.md
# M&A Deal Workbench — Structured Deal Review Skill
## Overview
This skill enables you to act as deal counsel reviewing an M&A transaction using a running **M&A Deal Workbench** REST API. You gather deal records, draft terms, playbook/policy rules, benchmarks, risk estimates, and related diligence data, then produce a structured JSON answer conforming to a task-specific answer template.
## Activation
Invoke this skill when the task involves:
- A deal ID (e.g. `PRJ_*`)
- An M&A deal workbench base URL (provided as `<TASK_ENV_BASE_URL>`)
- An answer template JSON (provided as `input/payloads/answer_template.json`)
- A playbook ID or policy ID for comparison
- Any of these review types: issue register, closing package, escalation memo, transition review, or deviation matrix
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,000 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Available Endpoints
- Systematic Workflow
- Output Types & Patterns
- Enum Reference
- Common Pitfalls
Files
SKILL.mdcore13,533 bytesSKILL.mdcore13,533 bytes
Read SKILL.md
# M&A Deal Workbench — Structured Deal Analysis
You are an M&A transaction counsel working inside a deal workbench environment.
Your job is to gather deal records, compare draft terms against an applicable
playbook or policy, quantify exposures, and produce a structured JSON analysis.
---
## Environment
| Setting | Value |
|--------------------|----------------------------------------------|
| Base URL | `<TASK_ENV_BASE_URL>` |
| Read-only SQL token | `deal-workbench-readonly` |
| Output language | English only |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,166 tokensWhen the skill is activated
Reference documents
1,405 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Workflow
- Deal-Type Specific Guidance
- Common Pitfalls
Files
SKILL.mdcore9,916 bytesworkbench_reference.mddoc5,754 bytes
Read SKILL.md
# M&A Deal Workbench — Transaction Counsel Skill
You are an M&A transaction counsel agent. When given a task prompt and an answer template, use the running M&A deal workbench to gather deal records, apply playbook or policy rules, compute quantified positions, and return a single JSON answer conforming exactly to the template.
---
## Workflow
### 1. Read the Prompt and Template
Extract from the prompt:
- **deal_id** — the project code (e.g. PRJ_ALPHA)
- **client_side** — `seller` or `buyer`
- **transaction type** — asset purchase, stock purchase, merger, or carveout
- **deliverable type** — issue register, closing package, escalation memo, transition review, deviation matrixWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,072 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Core workflow
- Task-type quick reference
- Common pitfalls
Files
SKILL.mdcore17,606 bytesSKILL.mdcore17,606 bytes
Read SKILL.md
# M&A Deal Workbench — Counsel Task Skill
## When to use
This skill applies whenever you are given a structured M&A transaction counsel task
with a deal workbench that exposes deal records, draft terms, playbooks, policies,
and diligence data through HTTP APIs. The task will ask you to prepare a structured
JSON answer — an issue register, closing package, committee escalation memo,
transition review, or deviation matrix — by gathering facts from the workbench and
comparing them against a playbook or policy.
## Core workflow
### 1. Orient from the prompt
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,698 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Core Workflow
- Task-Type Variations
- Reference: Common Framework Rules
- Units and Precision (by convention across tasks)
Files
SKILL.mdcore11,622 bytesSKILL.mdcore11,622 bytesSKILL.mdcore11,622 bytes
Read SKILL.md
# M&A Deal Workbench Skill
## Purpose
Solve structured M&A transaction counsel tasks using a running deal workbench. The workbench serves deal records, draft terms, playbook rules, policy thresholds, and diligence data through a REST API and optional read-only SQL. Every task requires gathering data from the workbench, comparing draft positions against a governing framework (playbook or policy), producing a structured JSON answer that conforms to a supplied answer template, and returning only that JSON.
## Core Workflow
Follow this sequence for every task; skip steps only when the task scope explicitly excludes them.
### 1. Read the prompt and template together
- Extract `deal_id`, `client_side` (buyer or seller), and the governing framework (`playbook_id` or `policy_id`) from the prompt.
- Read `input/payloads/answer_template.json` to learn the exact output shape, allowed enums, stable IDs, required fields, units, and ordering rules.
- Note any task-specific filters: e.g., "only current terms," "exclude stale," "only out-of-policy," "transition/separation focus only."When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,803 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Phase 1 — Environment Bootstrap
- Phase 2 — Task Parsing
- Phase 3 — Load the Answer Template
- Phase 4 — Data Gathering
- Phase 5 — Analysis & Comparison
- Phase 6 — Quantification
- Phase 7 — Output Assembly
- Phase 8 — Validation (Before Returning)
- Endpoint Quick Reference
- Enum Reference
Files
SKILL.mdcore11,139 bytesSKILL.mdcore11,139 bytes
Read SKILL.md
# M&A Deal Workbench — Reusable Operating Rules
## Purpose
This skill governs interaction with the M&A deal workbench — a structured API surface for reviewing acquisition agreements (APA, SPA, carveout, or other M&A transaction types). Apply these rules whenever the task involves the deal workbench, regardless of client side (buyer or seller), transaction structure, or deliverable format.
---
## Phase 1 — Environment Bootstrap
1. **Read `environment_access.md`** from the working directory. Extract:
- `base_url` — the workbench root (e.g. `http://task-env:9020/`).
- `credentials.read_only_sql_token` — the bearer token for `POST /api/query`.
- `allowed_endpoints` — the full list of permitted routes. Do not call endpoints outside this list.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,223 tokensWhen the skill is activated
Reference documents
732 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Environment
- Available API endpoints
- Core operating rules
- Analysis type patterns
- Cross-cutting patterns
- Error handling and edge cases
Files
SKILL.mdcore14,677 bytesSKILL.mdcore14,677 bytesSKILL.mdcore14,677 bytesapi_reference.mddoc2,909 bytes
Read SKILL.md
# M&A Deal Workbench — Structured Analysis Skill
## When to use
Invoke this skill whenever the task involves analyzing M&A deal terms using the deal workbench API — producing issue registers, economics packages, escalation memos, transition reviews, deviation matrices, or any structured JSON output that compares draft deal terms against playbook rules, policy thresholds, benchmarks, or market standards.
## Environment
The deal workbench runs at a base URL provided as `<TASK_ENV_BASE_URL>`. All API routes are relative to this base. Read `environment_access.md` (in the task root) for the actual base URL and credentials — do not guess them.
- **Read-only SQL token**: `deal-workbench-readonly` — use with `POST /api/query` for cross-table or aggregate checks when the standard endpoints are insufficient.
- **Web UI**: The workbench also serves a browsable interface at the base URL. Use it for discovery when API responses are ambiguous.
## Available API endpoints
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,634 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Environment
- Operating Rules
- Task-Type-Specific Guidance
Files
SKILL.mdcore11,951 bytes
Read SKILL.md
# M&A Deal Workbench Skill
## Purpose
You are an M&A transaction counsel operating a deal workbench at a configurable base URL. Your task is to gather deal records, compare draft terms against playbook/policy positions, and produce structured JSON outputs following a supplied answer template. This skill covers seller-side and buyer-side representations across asset purchase agreements (APA), stock purchase agreements (SPA), and carveout transactions.
## Environment
The deal workbench is available at the base URL specified in the task's environment configuration (`environment_access.md` or `<TASK_ENV_BASE_URL>`). The workbench exposes both a web UI and a REST API.
### API Access
All GET endpoints return JSON. The allowed endpoints are listed in the environment configuration. Common endpoints include:
| Endpoint | Purpose |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,541 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Input files (every task provides these)
- Phase 1 — Connect and discover the data landscape
- Phase 2 — Acquire and reconcile the data
- Phase 3 — Quality and integrity checks
- Phase 4 — Compute normalized totals
- Phase 5 — Focus-item resolution
- Phase 6 — Ranking outputs
- Phase 7 — Internal control and policy codes
- Phase 8 — Certification / reconciliation status
- Phase 9 — Output assembly
- Phase 10 — Verification checklist
Files
SKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytesSKILL.mdcore15,531 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Certification Skill
## Purpose
Complete Asteria Fleet data-certification tasks by reconciling source records, applying quality rules,
assigning internal control codes, and producing a JSON answer that conforms exactly to the supplied
output contract (`answer_template.json`).
## Input files (every task provides these)
| File | Role |
|---|---|
| `prompt.txt` | Business narrative and list of required outputs. |
| `payloads/case_scope.json` | Task parameters: collection, cutoff, focus items, threshold rules, ranking policies. |
| `payloads/answer_template.json` | JSON Schema or field contract the output must satisfy. |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,864 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Environment Setup
- API Reference
- Standard Workflow
- Output Rules
- Domain-Specific Notes
- Common Pitfalls
Files
SKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytesSKILL.mdcore17,660 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Audit & Certification Skill
## Purpose
Reconcile overlapping source records from the Asteria Fleet Data Quality Hub, compute integrity metrics, assign internal control codes, and produce a certification decision. This skill covers auditing transactional collections (fuel, freight, maintenance) and certifying entity populations (contacts, partners). Every task is scoped by a `case_scope.json` payload and must produce a single JSON object conforming exactly to an `answer_template.json` contract.
## Environment Setup
Connection details are always supplied in a companion file `environment_access.md` containing:
- `base_url` — the root URL of the Fleet Data Quality Hub (e.g. `http://task-env:9021/`)
[Credential-related source omitted]
- `allowed_endpoints` — the list of API paths available for the task
- `runtime_notes` — any override of `<TASK_ENV_BASE_URL>` references found in task prompt text
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,447 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Input File Contract
- Environment Bootstrap
- Discovery & Exploration
- Reconciliation Methodology
- Quality & Integrity Metrics
- Exception Classification
- Focus-Entity Reporting
- Channel / Communication Readiness
- Control-Code Assignment
- Rankings
- Regional / Dimensional Rollups
- Certification / Close Decision
- Output Compliance
- Execution Checklist
Files
SKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytesSKILL.mdcore15,778 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Audit & Certification Skill
## Purpose
Reconcile, audit, and certify fleet-domain datasets published through the **Asteria Fleet Data Quality Hub** (a shared, read-only HTTP API). Produce a structured JSON certification report conforming to a supplied answer-template schema.
## When to Use
This skill applies whenever a task involves:
- An Asteria collection name, case-scope payload, and answer-template payload provided as input files (`prompt.txt`, `payloads/case_scope.json`, `payloads/answer_template.json`).
- Runtime connection details supplied in a separate `environment_access.md`.
- Instructions to query the hub, reconcile overlapping source records, compute quality/integrity metrics, assign opaque control codes, and render a certification or close decision.
## Input File Contract
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,779 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Step 1 — Understand the Data Landscape
- Step 2 — Examine Reference Data
- Step 3 — Reconcile Source Snapshots
- Step 4 — Classify Records
- Step 5 — Compute Aggregations
- Step 6 — Assign Decision Codes
- Step 7 — Apply Certification Thresholds
- Step 8 — Assemble the Answer
- Key API Conventions
- Self-Check Before Submitting
Files
SKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytesSKILL.mdcore12,449 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Reconciliation Skill
## Purpose
Use the Asteria Fleet Data Quality Hub REST API to audit, reconcile, classify, and normalize multi-source fleet records for a business-scoped collection, producing a structured answer that conforms exactly to a supplied answer-template contract.
## Step 1 — Understand the Data Landscape
### 1.1 Discover available collections
```
GET /api/catalog/collections
```
Returns every available collection with its `collection_id`, `family`, `source_systems`, `time_start`/`time_end`, and approximate row count. Match the `collection_id` from the task's case scope (`payloads/case_scope.json`) to confirm it exists.
### 1.2 Inspect the logical schema
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,552 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Step 1 — Discover the collection and its schema
- Step 2 — Load source-snapshot metadata
- Step 3 — Load reference data
- Step 4 — Load and deduplicate the primary data
- Step 5 — Detect and classify issues
- Step 6 — Reconcile canonical entities (contacts only)
- Step 7 — Assign control codes
- Step 8 — Compute normalised totals (transactions only)
- Step 9 — Channel readiness (contacts only)
- Step 10 — Compute rankings
- Step 11 — Determine certification / reconciliation status
- Step 12 — Format and validate the answer
- Quick-reference: source-system precedence for contacts
- Quick-reference: normalisation functions
- Quick-reference: snapshot precedence
Files
SKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytesSKILL.mdcore15,054 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Audit & Certification Skill
## Overview
Use this skill to process Asteria Fleet data-quality audit and certification tasks. Every task reconciles overlapping source records from a shared read-only data hub, detects data-quality issues, normalises values against reference tables, assigns internal control codes, and returns a single JSON object conforming to a supplied answer contract.
## Step 1 — Discover the collection and its schema
Read the task's `case_scope.json` for the `collection_id` and business cutoff or as-of date. Read `environment_access.md` for the base URL, authorization header, and the list of allowed endpoints.
**Catalog.** Call `GET /api/catalog/collections` to confirm the collection exists and to learn its source systems, approximate row count, and time range.
**Schema.** Call `GET /api/catalog/schema` to list every available logical view and its fields. Match the task domain — contacts, fuel transactions, freight charges, maintenance events — to the correct view (`v_contacts`, `v_fuel_transactions`, `v_freight_charges`, `v_maintenance_events`, `v_reference_aliases`, `v_unit_conversions`, `v_fx_rates`, `v_source_snapshots`). Study the `meaning` annotations on each field; they describe the business purpose and expose aliases (e.g., `purchased_description` in fuel is the text you match against reference aliases).
## Step 2 — Load source-snapshot metadataWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,192 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Phase 1 — Understand the Task Scope
- Phase 2 — Explore the Collection
- Phase 3 — Deduplicate and Reconcile
- Phase 4 — Normalize and Classify
- Phase 5 — Compute Aggregates
- Phase 6 — Assign Internal Control Codes
- Phase 7 — Certify
- Quick Reference: Common Pitfalls
Files
SKILL.mdcore9,515 bytesSKILL.mdcore9,515 bytesSKILL.mdcore9,515 bytesSKILL.mdcore9,515 bytesSKILL.mdcore9,515 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Reusable Reconciliation Skill
## Overview
When you receive a data-quality certification or audit task against the Asteria Fleet Data Quality Hub, follow the systematic process below. The hub exposes a read-only SQL query interface, reference tables, and collection metadata. Every task follows the same reconciliation pattern: **scope → deduplicate → normalize → classify → certify**.
---
## Phase 1 — Understand the Task Scope
1. **Read the case scope** (`payloads/case_scope.json`) for:
- `collection_id` — which data collection to audit
- `cutoff_at` / `as_of` — business-time cutoff; only rows with `business_updated_at` ≤ cutoff are in scope
- `focus_*` lists — specific entities (assets, people, transactions) requiring detailed per-row decisions
- Thresholds and ranking policies — determine the final certification statusWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,196 tokensWhen the skill is activated
Reference documents
752 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Input Convention
- API Interaction
- Reconciliation Rules
- Numeric Precision
- Output Rules
- Certification / Release Decision
- Verification Checklist
Files
SKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesSKILL.mdcore10,357 bytesreference_codes.mddoc3,769 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Reconciliation Skill
## Purpose
Execute a data-quality reconciliation, audit, or certification task against the Asteria Fleet Data Quality Hub. The hub exposes a read-only REST API over fleet-domain collections (contacts, fuel transactions, freight charges, maintenance events). Each task reconciles overlapping source records, applies business rules, assigns internal control codes, and produces a single JSON answer conforming to a provided output contract.
## When to Use
Invoke this skill when the user prompt references:
- "Asteria Fleet Data Quality Hub" or a `<TASK_ENV_BASE_URL>` placeholder
- Reconciling fleet-domain records across source snapshots
- Producing a certification, audit, or readiness decision with control codes
- An `environment_access.md` file with hub connection details
- `payloads/case_scope.json` and `payloads/answer_template.json`
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,579 tokensWhen the skill is activated
Reference documents
1,425 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Trigger
- Environment Setup
- Input Contract
- Operating Procedure
- Output Rules
- Certification Matrix
- Hub API Reference
Files
SKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesSKILL.mdcore16,289 bytesreference/codes.mddoc6,129 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Audit & Reconciliation Skill
## Purpose
Execute data-quality audit, entity reconciliation, canonical-master certification, and accrual-close tasks against the Asteria Fleet Data Quality Hub. This skill covers any task that: (a) involves the Asteria Fleet Data Quality Hub as the data source, (b) requires reconciling overlapping source records into canonical entities, (c) assigns opaque internal control/decision codes derived from business evidence, and (d) produces a certification or release decision from business thresholds.
## Trigger
Invoke this skill when the task prompt references the **Asteria Fleet Data Quality Hub**, instructs you to conform to an `answer_template.json` or `payloads/answer_template.json`, includes a `payloads/case_scope.json`, and/or mentions operations such as: contact-master certification, fuel-purchase audit, maintenance-log integrity certification, field-service roster readiness, or freight-charge accrual reconciliation.
## Environment Setup
Before any work, read `environment_access.md` at the task root. This file supplies:
- `base_url` — the actual origin for all API calls (overrides `<TASK_ENV_BASE_URL>`, `localhost`, `127.0.0.1`, and any `TASK_ENV_BASE_URL` env var)asteria-fleet-data-quality-hub
When your agent uses it
Reusable operating rules for reconciling and certifying data-quality collections against the Asteria Fleet Data Quality Hub REST API.
What it loads into context
Frontmatter
40 tokensAlways available to the agent
SKILL.md
3,492 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Input Files
- Discovery Workflow
- Reconciliation Patterns
- Control Codes
- Thresholds and Certification
- Normalized Totals
- Ranking
- Output Rules
- Endpoint Reference
- Task Execution Sequence
- Common Pitfalls
Files
SKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytesSKILL.mdcore16,374 bytes
Read SKILL.md
# Asteria Fleet Data Quality Hub — Reusable Operating Rules
## Environment
The Asteria Fleet Data Quality Hub is a read-only REST API. Runtime access details are supplied in `environment_access.md`:
- **Base URL**: specified in `environment_access.md` under `base_url`
- **Auth header**: specified under `credentials.query_authorization_header`
- **Allowed endpoints**: listed under `allowed_endpoints`
- **Query endpoint**: `POST /api/query` accepts `{"query": "<SQL>"}` with the auth header
Apply the base URL, auth header, and endpoint list from `environment_access.md` for every call. Do not use localhost, 127.0.0.1, or any other base URL.
## Input FilesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,646 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Input layout (every task)
- Workplace API
- Step-by-step method
- Task-type quick reference
- Guardrails
Files
SKILL.mdcore11,939 bytesSKILL.mdcore11,939 bytesSKILL.mdcore11,939 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical & Transactional Skill
## Purpose
Complete a self-contained operational data task against the Atlas Commerce Operations
workplace. Each task provides a natural-language prompt, a business-request payload with
scope/definitions/rules, and an answer-template JSON that enforces the output schema.
Produce exactly one `answer.json` that matches the template with no extra keys or
narrative.
## Input layout (every task)
```
input/
prompt.txt — natural-language task brief + constraintsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,196 tokensWhen the skill is activated
Reference documents
466 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Phase 1 — Connect
- Phase 2 — Orient
- Phase 3 — Explore the data model
- Phase 4 — Query with SQL
- Phase 5 — Compute results
- Phase 6 — Write the answer
- Phase 7 — Data correction (only when requested)
- Common pitfalls
Files
SKILL.mdcore9,801 bytesSKILL.mdcore9,801 bytesSKILL.mdcore9,801 bytesapi_reference.mddoc1,757 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical Task Runner
## Purpose
Execute structured business-analytics tasks against the Atlas Commerce Operations
database. Each task supplies a natural-language prompt, a request payload with
business scope and definitions, and an output contract; this skill produces the
correct `answer.json`.
---
## Phase 1 — Connect
Read the file `environment_access.md` (at the workspace root or the nearest
ancestor directory). It contains three mandatory fields:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,456 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Environment
- Workflow
- API reference
- Cautions
Files
SKILL.mdcore6,709 bytesSKILL.mdcore6,709 bytesSKILL.mdcore6,709 bytesSKILL.mdcore6,709 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical & Data-Correction Tasks
## When to use
Invoke this skill whenever a task involves:
- Querying an **Atlas Commerce Operations** database through a REST API
- Fulfillment, refund, carrier, warehouse, or support-case analytics
- A prompt that references `<TASK_ENV_BASE_URL>`, Atlas, or input `payloads/*.json` files
- Read-only analytical scorecards **or** controlled data corrections (mutations) with audit verification
## Environment
Every run receives its target environment through a placeholder string (e.g. `<TASK_ENV_BASE_URL>`) in the task prompt. The actual base URL and credentials are provided in an `environment_access.md` file in the working directory.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,618 tokensWhen the skill is activated
Reference documents
467 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Entry workflow
- Common traps
- Appendix A: Effective-final determination
- Appendix B: Money and FX conversion
- Appendix C: Data correction workflow
Files
SKILL.mdcore11,578 bytesSKILL.mdcore11,578 bytesSKILL.mdcore11,578 bytesSKILL.mdcore11,578 bytesSKILL.mdcore11,578 bytesdedup_pattern.mddoc1,847 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical Task Skill
## When to use
Use this skill when working with the Atlas Commerce Operations workplace database to answer analytical business questions. The environment provides a read-only SQL API, a controlled transaction API for data corrections, an audit view, and business documents that define cohorts, metrics, thresholds, and output schemas.
---
## Entry workflow
### 1. Explore the schema and data dictionary first
Before writing any analytical query, fetch the full schema and data dictionary from the authenticated workplace service:
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,362 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to Use
- Core Environment
- Deduplication (Universal Rule)
- Effective Final State (Append‑Only Event Tables)
- Money and FX
- Correction Workflow
- Query Strategy
- General Workflow
Files
SKILL.mdcore6,159 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical Task Skill
## When to Use
Invoke this skill when asked to produce a data-analysis answer from the Atlas Commerce Operations workplace: scorecards, reconciliations, quality reviews, productivity reports, or health reviews that require SQL queries against the provided database and submission of a structured JSON answer matching a supplied template.
## Core Environment
[Credential-related source omitted]
- `GET /api/schema` — table DDL and indexes.
- `GET /api/data-dictionary` — column descriptions, conventions (timestamps are ISO‑8601 UTC, money uses smallest currency units, raw vs canonical fields).
- `POST /api/sql` — read‑only analytical queries. Body: `{"sql": "<SQL>", "params": []}`.
- `GET /api/correction-audit` — lists applied canonical corrections.
- `POST /api/sql/transaction` — controlled writes (see Correction Workflow below).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,014 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Workflow
- Common patterns by domain
- Data quality conventions
Files
SKILL.mdcore8,642 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical Task Skill
## Purpose
Solve operational analytics and data-correction tasks against the Atlas Commerce Operations database. Each task provides a business-request payload and an answer template; produce a JSON answer conforming exactly to the template.
## Workflow
### Phase 1: Orient
1. **Read the schema** (`GET /api/schema`). Identify all tables that participate in the task. Note primary keys, foreign keys, and check constraints — they encode domain invariants.
2. **Read the data dictionary** (`GET /api/data-dictionary`). Note field-level conventions: timestamp format (ISO-8601 UTC), monetary minor units, boolean encoding (0/1 integers), and dedup indexes.
3. **Read the task prompt** and the **business-request payload**. Extract: cohort/scope rules, date windows, cutoff timestamps, business definitions, rollup instructions, rounding rules, and status-tier conditions.
4. **Read the answer template**. Every `required` field must appear; every `additionalProperties: false` constraint means no extra fields. Match `type`, `enum`, `pattern`, `minimum`/`maximum`, `multipleOf`, and `minItems`/`maxItems` exactly.
### Phase 2: Scope the cohort
- **Production accounts**: `is_internal = 0 AND is_test = 0`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,143 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Step 0 — Locate the task inputs
- Step 1 — Read environment access
- Step 2 — Discover the data model
- Step 3 — Parse the request-facts JSON
- Step 4 — Query the data (read-only analysis)
- Step 5 — Apply corrections (only when the task requires mutation)
- Step 6 — Write the conformant output
- Reference — API summary
- Reference — common business-logic patterns
- Reference — error recovery
Files
SKILL.mdcore13,531 bytesSKILL.mdcore13,531 bytesSKILL.mdcore13,531 bytesSKILL.mdcore13,531 bytesSKILL.mdcore13,531 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical & Correction Skill
## When to use
Invoke this skill whenever a task requires querying or correcting data in the Atlas Commerce Operations database through the authenticated workplace API. The task will involve reading structured business-scope payloads, running SQL against the service, applying domain policies (classification, ranking, aggregation), and returning a strictly conformant JSON answer.
## Step 0 — Locate the task inputs
Every task instance lives in a directory with this shape:
```
<input_dir>/
prompt.txt # natural-language summary of the business request
payloads/
<request_facts>.json # structured scope, policies, thresholds, business definitionsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,591 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Invocation
- Environment Resolution
- API Reference
- Task Execution Pipeline
- Business Rule Patterns
- Error Recovery
Files
SKILL.mdcore12,211 bytesSKILL.mdcore12,211 bytesSKILL.mdcore12,211 bytesSKILL.mdcore12,211 bytes
Read SKILL.md
# Atlas Commerce Operations — Analytical Task Runner
## Purpose
Execute analytical and corrective tasks against the Atlas Commerce Operations database through its authenticated HTTP API. Each task follows a standard discovery→query→compute→output pipeline, producing a single `answer.json` that conforms to a supplied output schema.
## Invocation
This skill applies when a task directory contains:
- `input/prompt.txt` — natural-language task description referencing `<TASK_ENV_BASE_URL>`
- `input/payloads/*.json` — one or more JSON payloads with business scope, definitions, and rules
- `input/payloads/answer_template.json` — the exact output JSON schema
The skill is also triggered by references to "Atlas Commerce Operations", "Atlas workplace", or the endpoint names `/api/schema`, `/api/data-dictionary`, `/api/sql`, `/api/sql/transaction`, `/api/correction-audit`.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,009 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Preconditions
- API Reference
- Workflow
- Cross-Cutting Rules
Files
SKILL.mdcore9,235 bytesSKILL.mdcore9,235 bytesSKILL.mdcore9,235 bytes
Read SKILL.md
# Atlas Operations Skill
Reusable skill for analyzing and correcting business data through the Atlas Commerce Operations authenticated workplace API. Handles scorecards, reconciliations, quality corrections, productivity reviews, and support-health assessments.
## Preconditions
Before executing any task, confirm these files are present in the working directory:
- `input/prompt.txt` — the task brief
- `input/payloads/<request_file>.json` — business scope, definitions, policies, and rules
- `input/payloads/answer_template.json` — output contract (schema, required fields, constraints)
- `environment_access.md` — base URL, credentials, and allowed endpoints (overrides any `<TASK_ENV_BASE_URL>` placeholder in the prompt)
If any file is missing or the working directory contains material not matching this structure, stop and report the discrepancy before proceeding.
## API ReferenceWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,644 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Portal Reference
- The Analysis Contract
- Protocol Registry and Override Resolution
- Reusable Algorithmic Modules
- Answer Template Contract
- Answer Assembly Rules
- Execution Checklist
Files
SKILL.mdcore29,159 bytesSKILL.mdcore29,159 bytesSKILL.mdcore29,159 bytesSKILL.mdcore29,159 bytesSKILL.mdcore29,159 bytes
Read SKILL.md
# Public Health Observatory — Algorithmic Transport Audit Skill
## Overview
This skill guides an agent through completing a registered algorithmic audit against the Public Health Observatory (PHO) read-only evidence portal. Every audit is defined by an `analysis_request.json` (the effective request) and an `answer_template.json` (the response contract). The agent resolves evidence from the portal, executes each declared module in order, and returns exactly one JSON object conforming to the answer template — no narrative outside the JSON.
## Portal Reference
The portal is reached at the base URL supplied as `<TASK_ENV_BASE_URL>`. All endpoints are read-only GET with no authentication required.
### Available Endpoints
| Endpoint | Purpose |
|---|---|
| `GET /` | Portal root / health check |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,098 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Input Artifacts
- Portal Evidence Access
- Reusable Audit Module Methods
- Controlled Decision
- Output Construction
- Execution Checklist
Files
SKILL.mdcore20,295 bytesSKILL.mdcore20,295 bytesSKILL.mdcore20,295 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit Skill
Activate when a task references a Public Health Observatory (`PHO_*`) protocol, an `analysis_request.json` paired with an `answer_template.json`, or a registered algorithmic audit over health/socioeconomic evidence served from a read-only Observatory web portal.
## Overview
This skill executes registered algorithmic audits for the Public Health Observatory. Every audit follows the same pattern:
1. Read the three input artifacts: the task prompt, `analysis_request.json`, and `answer_template.json`.
2. Access the Observatory web portal at the base URL specified in the task (typically the `<TASK_ENV_BASE_URL>` placeholder) using only the allowed endpoints.
3. Resolve evidence — publication records, releases, revisions — according to the effective request's filters.
4. Execute each declared audit module in the registered order, applying the exact method specification.
5. Construct a single JSON output object conforming to the answer template, preserving every declared order, identifier, precision, and type.
6. Evaluate the controlled decision gates and return the mandated classification.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,174 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Portal Access
- Override Resolution: Building the Effective Request
- Module Execution Order
- Reusable Module Methods
- Controlled Decision
- Reporting Conventions
- Execution Checklist
- Portal Interaction Notes
Files
SKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytesSKILL.mdcore22,158 bytes
Read SKILL.md
# PHO Algorithmic Audit Skill
## Purpose
Execute Public Health Observatory (PHO) registered algorithmic audits. Given an
`analysis_request.json` and an `answer_template.json`, access the read-only PHO
data portal, resolve publication records, construct analytic cohorts, run every
declared audit module in its registered order, evaluate controlled business
predicates, and return one completed JSON object conforming to the template.
This skill describes the reusable method. Every entity identifier, measure code,
time coordinate, source filter, random seed, hyperparameter grid value, business
cutoff, and output label is drawn from the effective request — never carried
across invocations.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,971 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Portal Access
- Data Pipeline
- Cohort Construction
- Statistical Computation (Pure Python / JS)
- Output Formatting
- Common Pitfalls
- Module Patterns
- Execution Workflow
- Self-Check Before Submission
- Data-Only Modules (Highest Leverage)
Files
SKILL.mdcore13,082 bytesSKILL.mdcore13,082 bytesSKILL.mdcore13,082 bytesSKILL.mdcore13,082 bytes
Read SKILL.md
# Public Health Observatory Audit Skill
## Overview
Solve algorithmic audit tasks against the Public Health Observatory data portal. Each task asks you to complete a registered multi-module evidence audit using only the read-only portal. You must return one JSON object conforming exactly to the supplied answer template.
## Portal Access
The portal is a read-only Web application. All evidence comes from browse pages or CSV exports. Never modify data — every filter, release-resolution step, and quality exclusion must be reproducible from the portal alone.
### Data Endpoints
| Endpoint | Returns |
|---|---|
| `GET /geographies/states` | State reference: FIPS, abbreviation, name, region, census division |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,792 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Portal Navigation
- Data Resolution Rules
- Reading analysis_request.json
- Reading answer_template.json
- Implementation Approach
- Supporting Files
Scripts executed separately
linalg.py
Files
SKILL.mdcore20,913 bytesSKILL.mdcore20,913 bytesSKILL.mdcore20,913 byteslinalg.pyruns17,352 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit Skill
## Overview
You are completing a registered algorithmic audit for the Public Health Observatory (PHO). You receive three inputs:
1. **prompt.txt** — the business question and scope
2. **analysis_request.json** — the registered protocol specifying statistical modules, parameters, cohorts, decision rules, and numerical precision
3. **answer_template.json** — the exact JSON output contract with required keys, types, array lengths, and ordering constraints
Your job is to fetch data from the Observatory's read-only web portal, perform every declared computation, and return a single JSON object conforming exactly to the answer template.
## Portal Navigation
The portal is a read-only HTML + CSV data service. The base URL is provided by the task environment (look for `<TASK_ENV_BASE_URL>` or similar in prompt.txt).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,936 tokensWhen the skill is activated
Reference documents
618 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Core workflow
- Portal reference
- Common pitfalls
Files
SKILL.mdcore17,151 bytesSKILL.mdcore17,151 bytesSKILL.mdcore17,151 bytesSKILL.mdcore17,151 bytesSKILL.mdcore17,151 bytesSKILL.mdcore17,151 bytesgeography_reference.mddoc2,258 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit Skill
## When to use
Invoke this skill when the task involves:
- A Public Health Observatory (PHO) algorithmic audit, transportability assessment, or publication briefing
- An `analysis_request.json` paired with an `answer_template.json` defining a registered audit protocol
- State-level, county-level, or country-level health/socioeconomic data from the PHO read-only web portal
- Statistical modules including: fixed-effects OLS with jackknife, nested ridge/elastic-net cross-validation, wild cluster bootstrap, grouped split conformal calibration, trajectory PCA clustering, source perturbation analysis, GMM mediation, or sensitivity surfaces
## Core workflow
### Phase 1: Understand the contract
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,015 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- When to Use
- Prerequisites
- Workflow
- Cross-Cutting Operating Rules
- Portal Endpoint Reference
- Task Execution Checklist
Files
SKILL.mdcore13,913 bytesSKILL.mdcore13,913 bytesSKILL.mdcore13,913 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit Skill
## Purpose
Execute a registered multi-module algorithmic audit using the Public Health Observatory (PHO) read-only web portal. Produce a single JSON response conforming exactly to a provided answer template — no narrative outside the JSON.
## When to Use
Invoke this skill when the task involves:
- A PHO analysis request (`analysis_request.json`) with declared audit modules
- A companion answer template (`answer_template.json`) specifying the output contract
- A read-only web portal exposing health, socioeconomic, and geographic data
## Prerequisites
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,470 tokensWhen the skill is activated
Reference documents
1,279 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Trigger
- Portal
- Workflow
- Audit module reference
- Decision rule patterns
- Quality flags and missing data
- Precision reference
Files
SKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesSKILL.mdcore21,332 bytesapi_reference.mddoc5,076 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit Skill
## Trigger
Use this skill when the task involves completing a registered algorithmic audit for the Public Health Observatory (PHO) — a structured, multi-module reproducibility and transportability assessment of a public-health association — using the PHO read-only web portal as the sole evidence source.
**Detect via:** presence of `analysis_request.json` and `answer_template.json` payloads, references to a `<TASK_ENV_BASE_URL>` portal, or terms like "registered audit," "algorithmic audit," "transportability audit," "publication gates," or "robustness decision."
## Portal
| Setting | Value |
|---------|-------|
| Base URL | `http://task-env:9023/` |
| Auth | None required |
| Protocol | HTTP GET only |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,691 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Trigger
- Portal Interaction
- Data Selection and Cohort Construction
- Statistical Methodology Patterns
- Output Formatting Rules
- Decision Rule Application
- Verification Checklist
Files
SKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytesSKILL.mdcore17,669 bytes
Read SKILL.md
# Public Health Observatory Algorithmic Audit
## Trigger
Use this skill when the user needs to complete a Public Health Observatory (PHO) registered algorithmic audit — a structured, multi-module statistical analysis that queries the PHO read-only data portal, applies registered methodologies, and returns a single JSON object conforming to a declared answer template.
Trigger signals: the task references a PHO portal at a configurable base URL, includes an `analysis_request.json` with audit modules and decision rules, and provides an `answer_template.json` that fixes the output contract.
## Portal Interaction
The PHO portal is an unauthenticated read-only HTTP API. Discover available endpoints from the accompanying environment description. All data is available via GET requests returning JSON arrays of records.
**Base workflow:**
1. Fetch `/` or `/catalog` to confirm available datasets.
2. Identify the relevant endpoint(s) from the analysis request's declared datasets, geographies, and measures.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,049 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment Connection
- API Reference
- Portfolio Category Classification
- Task Type 1: Portfolio Mix Review
- Task Type 2: SLA Aging Audit
- Task Type 3: Release Readiness Assessment
- Cross-Cutting Conventions
- Common Pitfalls
- Tools and Commands Cheat Sheet
- Output Format
Files
SKILL.mdcore24,613 bytesSKILL.mdcore24,613 bytesSKILL.mdcore24,613 bytes
Read SKILL.md
# Portfolio Engineering Environment — Reusable Skill
## Overview
This skill covers interacting with a shared portfolio-management REST API to produce:
- **Portfolio mix reviews** — classify closed work items into investment categories, compare actual mix against targets, identify gaps, and recommend rebalancing actions.
- **SLA aging audits** — identify primary SLA-eligible work, detect overdue items, compute aging distributions, surface owner/team hotspots, flag duplicate clusters, and calculate breach rates.
- **Release readiness assessments** — evaluate milestone completion, identify gating work items, count unresolved high-impact blockers, trace critical dependency chains, and compute readiness scores.
---
## Environment Connection
### Base URL
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,141 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Read the Input Files
- 2. Connect to the Environment
- 3. Task Type Reference
- 4. General Data Conventions
- 5. Querying with POST /api/query
- 6. Output Requirements
- 7. Common Pitfalls
Files
SKILL.mdcore17,717 bytesSKILL.mdcore17,717 bytesSKILL.mdcore17,717 bytes
Read SKILL.md
# Portfolio Engineering Review Skill
You are an agent that completes portfolio engineering review tasks by interacting with a shared HTTP API environment. Follow these instructions to read the task prompt and answer template, query the environment, compute results, and produce a correctly formatted JSON answer.
---
## 1. Read the Input Files
Each task directory contains:
- `prompt.txt` — the task instructions, including scope parameters (teams, quarter, release id, as-of date, etc.)
- `payloads/answer_template.json` — the required JSON output schema
Read both files. The `prompt.txt` defines the business scope. The `answer_template.json` is a JSON Schema that defines every required field, its type, enum constraints, and ordering/naming conventions. **Your output must validate against this schema exactly.**
---When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,867 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- Environment
- Domain Model
- Common Conventions
- Workflow 1: Portfolio Mix Review
- Workflow 2: SLA Aging Analysis
- Workflow 3: Release Readiness Assessment
- Using POST /api/query
- Error Handling and Edge Cases
- Summary Checklist
Files
SKILL.mdcore17,213 bytesSKILL.mdcore17,213 bytes
Read SKILL.md
# Portfolio Engineering Operations Skill
## Overview
This skill provides reusable instructions for operating against a portfolio engineering REST API that exposes work items, mix targets, releases, milestones, SLA policies, blockers, and dependencies. It covers three common analytical workflows: **portfolio mix review**, **SLA aging analysis**, and **release readiness assessment**.
## Environment
The environment is accessed at `<TASK_ENV_BASE_URL>`. All endpoints are relative to this base URL.
### Authentication
For `POST /api/query` only, send the header:
```When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,826 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Authoritative Fields
- Work Item Classification
- Detecting Primary vs. Duplicate Records
- Scope and Filtering
- Computing SLA Metrics
- Computing Release Readiness
- Computing Portfolio Mix
- Ordering Conventions
- Precision
- General Guidance
Files
SKILL.mdcore12,468 bytes
Read SKILL.md
# Portfolio Engineering Skill
You are an agent that analyzes a shared engineering-portfolio REST API to answer structured JSON questions about portfolio mix, SLA aging, and release readiness. Follow the procedures below for every task.
## Environment
The API base URL is provided in the task prompt as `<TASK_ENV_BASE_URL>`. A companion file (`environment_access.md`) lists the available endpoints and any required authentication tokens. Read that file at the start of every task; its contents may differ across task instances.
All data retrieval uses these endpoints (names stable; exact URL paths may vary — always read `environment_access.md`):
- `GET /api/work-items` and `GET /api/work-items/{id}` — the full work item catalogue
- `GET /api/mix-targets` — target portfolio mix percentages by scope
- `GET /api/sla-policy` — severity-to-days mapping
- `GET /api/releases` and `GET /api/releases/{id}` — release metadata, blockers, and milestones
- `GET /api/milestones` — milestone definitionsWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,276 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Entry Checklist
- Environment and Data Sources
- Portfolio Category Classification
- Task Type 1 — Portfolio Mix Review
- Task Type 2 — SLA Aging Audit
- Task Type 3 — Release Readiness Assessment
- Validation Before Submitting
- Common Pitfalls
Files
SKILL.mdcore13,817 bytes
Read SKILL.md
# Portfolio Engineering Skill
Solve portfolio-mix, SLA-aging, and release-readiness assessment tasks against a shared API environment containing work items, targets, policies, releases, milestones, blockers, and dependencies.
## Entry Checklist
Before building an answer, read these three inputs:
1. **Task prompt** (`input/prompt.txt`) — scope, domain, target date or quarter, the specific question to answer.
2. **Answer template** (`input/payloads/answer_template.json`) — the exact JSON schema the answer must match. Every `required` field must be present. Every `enum` or `const` must be respected. Sort orders, rounding rules, and field descriptions in the schema are binding.
3. **Environment access** (`environment_access.md` at the project root) — base URL, available endpoints, any required headers or tokens.
## Environment and Data Sources
All tasks share a REST API at the base URL from the environment access file. Use only the endpoints listed there. The data model includes these entities:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,512 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment Access
- Portfolio Category Classification
- Work-Item Filtering
- SLA Aging Calculations
- Release Readiness
- Mix Table Computations
- Sorting and Precision Conventions
- Common Pitfalls
Files
SKILL.mdcore10,478 bytes
Read SKILL.md
# Portfolio Engineering Analysis Skill
Use this skill whenever the task involves portfolio-mix analysis, SLA-aging audits, or release-readiness assessments against a shared engineering-work-item environment. It encodes the conventions learned from portfolio-review and SLA-audit tasks so you can classify work items, compute mix metrics, surface overdue work, and assess release readiness without rediscovering the rules each time.
---
## Environment Access
The task environment exposes a set of REST endpoints through `{BASE_URL}`. The runtime supplies the base URL and any required tokens separately — do not hardcode them.
Available endpoints typically include:
- `GET /api/work-items` — paginated or full list of work items
- `GET /api/work-items/{item_id}` — single item detail
- `GET /api/mix-targets` — portfolio mix targets by scope
- `GET /api/sla-policy` — severity-to-days SLA mappingWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,309 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Purpose
- Trigger
- Environment Access
- Work Item Model
- Portfolio Mix Analysis
- SLA Aging Audit
- Release Readiness Assessment
- General Operating Rules
Files
SKILL.mdcore14,670 bytes
Read SKILL.md
# Portfolio Engineering Analysis Skill
## Purpose
Analyze portfolio engineering data — work items, mix targets, SLA policies, releases, milestones, blockers, and dependencies — to produce structured assessments: portfolio mix reviews, SLA aging audits, and release readiness reports.
## Trigger
Invoke this skill for portfolio mix analysis, SLA aging audits, release readiness assessments, or any task requiring classification and analysis of work items against portfolio targets and policies.
---
## Environment Access
The portfolio engineering environment exposes a REST API at a base URL provided via `environment_access.md` or `<TASK_ENV_BASE_URL>`. All endpoints live under this single base.
### AuthenticationWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
1,781 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment
- Data Integrity Rules
- Portfolio Categories
- Ordering Conventions
- Precision Standards
- Calculation Patterns
- Output Convention
- Workflow
Files
SKILL.mdcore8,058 bytesSKILL.mdcore8,058 bytes
Read SKILL.md
# Portfolio Work-Item Analysis Skill
Perform structured portfolio analysis against a shared API environment — mix reviews, SLA aging audits, and release-readiness assessments — using work items, targets, SLA policies, releases, milestones, blockers, and dependencies.
## Environment
The task environment provides a base URL (`<TASK_ENV_BASE_URL>`) with endpoints and credentials documented in `environment_access.md`. That file lists:
- The base URL for all API calls.
- An optional `X-Env-Token` header (name and value) required for `POST /api/query` (SQL endpoint).
- The set of allowed endpoints.
**Endpoint catalog** (availability varies per task):
| Method | Path | Purpose |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,156 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment Protocol
- 2. Work Item Data Model
- 3. Portfolio Category Classification
- 4. Portfolio Mix Analysis
- 5. SLA Aging Audit
- 6. Release Readiness Assessment
- 7. Ordering Conventions
- 8. Precision Rules
- 9. Workflow
- 10. Edge Cases and Defensive Rules
- 11. Template Compliance Checklist
Files
SKILL.mdcore17,696 bytesSKILL.mdcore17,696 bytes
Read SKILL.md
# Portfolio Engineering Audit Skill
Perform portfolio engineering audits across three analysis types: portfolio mix analysis,
SLA aging audits, and release readiness assessments. This skill encodes reusable operating
rules distilled from audit patterns observed in engineering portfolio environments.
---
## 1. Environment Protocol
### 1.1 Startup
Before any analysis, read the file `environment_access.md` in the working directory. It
provides the runtime base URL, credentials, and the list of allowed API endpoints. The
base URL overrides any localhost or `env/setup.sh` references found in task prompts.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,545 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & ground rules
- 1. Universal normalization conventions
- 2. Event-handoff conventions (sponsor + badge reconciliation)
- 3. Tradeshow prospecting conventions
- 4. Import-batch cleaning conventions
- 5. Field-by-field quick reference (most common schemas)
- 6. Common pitfalls
Files
SKILL.mdcore25,972 bytesSKILL.mdcore25,972 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Skill
A transferable playbook for solving HarborCRM post-event / prospecting / import-batch
handoff tasks. The remote dataset is deterministic and the business conventions below
are stable across tasks; only the event, tradeshow, or import batch varies.
## 0. Environment & ground rules
- Base URL (the only allowed API): `<remote-env-url>`
- All endpoints are read-only GET. Pipe `curl -s <url>` through `python3 -m json.tool`.
- Always read the task's own `answer_template.json` first. The template is the contract:
it dictates the exact top-level keys, item keys, **allowed enum values**, **ordering
rules**, and number formatting. Two tasks that look similar can use slightly different
enum vocabularies (e.g. `sensor_vendor_only` vs `sensor_only`) — always emit the
template's allowed values, never a synonym from memory.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,871 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & data access
- 1. Universal conventions
- 2. Family A — Event sponsor / lead handoff
- 3. Family B — Trade-show prospecting
- 4. Family C — Contact-import cleaning
- 5. Common pitfalls
Files
SKILL.mdcore23,379 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Skill
A transferable playbook for solving unseen HarborCRM tasks. HarborCRM is a shared
read-only CRM-marketing dataset exposed over a small JSON API. Tasks always end
with **one JSON object** conforming to a per-task `answer_template.json`.
The dataset seeds three task families. Identify which family a task belongs to,
then apply the matching rules below.
| Family | Trigger in prompt | Train pattern |
|---|---|---|
| **A. Event sponsor / lead handoff** | "event_id", "post-event", "reconcile", sponsor orders + badges + invoices | train_001 (simple), train_004 (detailed) |
| **B. Trade-show prospecting** | "show_id", "exhibitors", "meeting_interest", "campaign" | train_002 (simple), train_005 (ranked) |
| **C. Contact-import cleaning** | "import batch", "raw_contacts", "suppression", "clean" | train_003 |
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
9,473 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API Cheat Sheet
- 1. Identify the Task Type
- 2. Normalization Rules (Cross-Cutting — All Workflows)
- 3. Workflow A — Event Reconciliation
- 4. Workflow B — Tradeshow Prospecting
- 5. Workflow C — Import Batch Cleaning
- 6. Sorting Rules Reference
- 7. Enum Reference
- 8. Critical Pitfalls & Edge Cases
Files
SKILL.mdcore36,912 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Solver Skill
> **Purpose:** Teach a future solver to produce correct JSON for unseen HarborCRM test tasks.
> The environment is a read-only JSON API at `<remote-env-url>`.
> Always return ONE JSON object matching the task's `answer_template.json`. No prose outside JSON.
---
## 0. API Cheat Sheet
Base URL: `<remote-env-url>` (call with `curl -s`, pipe to `jq` or `python3 -m json.tool`)
| Endpoint | Used For |
|---|---|
| `GET /api/events` | List all events (key fields: `lead_opportunity_amount`, `followup_days_after_end`, `sponsor_followup_days_after_end`, `end_date`) |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,718 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Mindset & process
- 1. API access
- 2. Controlled enums (authoritative source: `/api/policies` + per-task template)
- 3. Cross-cutting normalization
- 4. Due-date arithmetic (CONFIRMED via judge)
- 5. Sponsor classification & revenue (event-handoff tasks)
- 6. Qualified non-sponsor leads (event-handoff tasks)
- 7. Badge reconciliation (edgeai-style tasks)
- 8. Tradeshow prospecting (train_002 / train_005 style)
- 9. Import-batch cleaning (train_003 style)
- 10. Things the judge taught us that are easy to get wrong
- 11. Verification recipe (for the solver at test time)
Files
SKILL.mdcore18,835 bytesSKILL.mdcore18,835 bytesSKILL.mdcore18,835 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Skill
A transferable skill for solving HarborCRM post-event / prospecting / import tasks. You are
given a remote read-only API, a train-task answer template, and a prompt. Produce ONE JSON
object matching the template. No prose outside the JSON.
## 0. Mindset & process
1. Read the prompt and the `answer_template.json` carefully. Identify the task CATEGORY
(event-sponsor-handoff, tradeshow-prospecting, import-batch-cleaning, badge-reconciliation).
2. Pull every endpoint the prompt names from the API (see §1). Derive values FROM THE API,
never from memory or the prompt examples.
3. The template may be either (a) a fill-in JSON or (b) a SCHEMA description (with
`description`/`required_top_level_keys`/`field_definitions`). In the schema case, emit ONLY
the data keys declared in `required_top_level_keys` — do NOT echo the schema metadata.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,666 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment
- 1. Universal output rules
- 2. Event reconciliation tasks (post-event sponsor + lead handoff)
- 3. Badge-level reconciliation variant (per-badge decisions + campaign members)
- 4. Prospecting tasks (tradeshow exhibitor qualification)
- 5. Import-batch cleaning tasks
- 6. Solve procedure per task
Files
SKILL.mdcore15,199 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Skill
Solve HarborCRM post-event / prospecting / import-batch tasks against the shared
read-only HarborCRM API. Produce ONE JSON object matching the task's
`answer_template.json` — exactly its declared fields, nothing extra, no prose.
## 0. Environment
- Base URL: `<remote-env-url>` (use as the ONLY data source).
- Call with `curl -s <BASE><endpoint> | python3 -m json.tool`.
- Dataset is deterministic (seed 41001). All endpoints return JSON.
### Public GET endpoints
- `GET /api/policies` — authoritative rule constants (platform enums, sponsor status enums).
- `GET /api/events` and `GET /api/events/{event_id}`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,107 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment
- 1. Data model (endpoints you will need)
- 2. Step-by-step solver workflow
- 3. Universal normalization & formatting rules
- 4. Pitfalls learned from judge feedback (concrete corrections)
- 5. Open uncertainties (treat carefully on test tasks)
Files
SKILL.mdcore17,644 bytesSKILL.mdcore17,644 bytesSKILL.mdcore17,644 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Solver
Transferable skill for solving HarborCRM post-event / prospecting / import-batch tasks.
Learned by iterating train tasks against the train-only judge. Generalized — no hardcoded
account names, ids, or numeric answers.
## 0. Environment
- Base API URL: supplied by the runner (e.g. `<remote-env-url>`). All data is read-only GET.
- Authoritative rule constants live at `GET /api/policies`:
- `prospecting.platform_enums` = `["AUV","ROV","Underwater Camera"]`
- `sponsor_handoff.status_enums` = `["paid_deferred","open_invoice","proposal_only","not_sponsor"]`
- Every task returns ONE JSON object only (no prose outside JSON), matching the provided `answer_template.json`.
- Re-read the answer template for EACH task: field names, enums, and ordering rules differ per task
even within the same family. Do not assume a field from one task exists in another.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,117 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment & API
- 2. Cross-cutting normalization rules (apply everywhere)
- 3. Task archetypes
- 4. Archetype A — Event handoff/reconciliation
- 5. Archetype B — Trade-show exhibitor prospecting
- 6. Archetype C — Import batch cleaning
- 7. Sorting rules (quick reference)
- 8. Output discipline
- 9. Cross-task consistency checks
Files
SKILL.mdcore24,910 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Task Skill
Self-contained playbook for solving HarborCRM post-event / prospecting / import tasks.
A solver needs only this document + the task prompt (which names an `event_id`,
`show_id`, or `batch_id`) + the remote API base URL.
## 1. Environment & API
- Base URL (read-only): `<remote-env-url>`
- All endpoints return JSON. Dataset is deterministic (seed 41001). Use `curl -s <url> | python3 -m json.tool`.
- NEVER call any judge/POST/scoring endpoint. Only the public GET endpoints below.
### Endpoints (and what they give you)
| Endpoint | Purpose |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,629 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API access
- 1. Authoritative constants — read `/api/policies` first
- 2. Cross-cutting conventions
- 3. Family A — Event handoff / reconciliation
- 4. Family B — Trade-show prospecting
- 5. Family C — Import cleaning
- 6. Opportunity / pipeline math summary
- 7. Sort-key reference
- 8. Execution checklist for any task
- 9. Reasoning notes & judgment calls (apply consistently)
Files
SKILL.mdcore22,632 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Handoff, Prospecting & Import Skill
Self-contained conventions for solving HarborCRM task-group tasks against the shared
read-only API. Three task families share one dataset and one set of conventions:
1. **Event handoff / reconciliation** — post-event CRM handoff for a completed event
(sponsor status reconciliation, badge→lead classification, opportunity totals,
finance follow-up, due dates, CRM action counts).
2. **Trade-show prospecting** — qualify exhibitors at a trade show into a CRM-ready
lead list (platform coverage, priority tiers, opportunity sizing, CRM
create/update decisions, ranking, exclusions).
3. **Import cleaning** — prepare a raw contact batch for CRM import (normalize,
de-duplicate, suppress, drop unusable, decide create/update, count campaign
members).
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,523 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- When to use
- Environment & API access
- Universal solver workflow
- Task archetypes → endpoints
- Decision table A — Sponsor status classification
- Decision table B — Badge exclusion classification & precedence
- Decision table C — CRM create-vs-update (accounts / contacts / campaign members)
- Decision table D — Trade-show prospecting qualification & exclusion
- Decision table E — Normalization transforms
- Decision table F — Duplicate resolution (import batches)
- Decision table G — Due-date arithmetic
- Decision table H — Pipeline & opportunity math
- Decision table I — Enums & allowed values
- Decision table J — Sorting
- Cross-task consistency rules
- Common pitfalls
Files
SKILL.mdcore22,749 bytesSKILL.mdcore22,749 bytesSKILL.mdcore22,749 bytes
Read SKILL.md
# HarborCRM CRM-Marketing Solver
Solve HarborCRM event/trade-show/import reconciliation tasks that return a single JSON object conforming to a provided `answer_template.json`. These tasks reconcile sponsor orders, finance invoices, badge scans, CRM accounts/contacts/opportunities, campaign members, trade-show exhibitors, meeting interest, and import batches into an import-ready CRM handoff.
## When to use
Use whenever a task asks you to: reconcile a HarborCRM event post-handoff, prepare a trade-show prospecting/campaign lead list, clean a CRM import batch, or produce sponsor-finance follow-up — and provides an output JSON schema (`answer_template.json`) plus a HarborCRM API. If a task instead asks for arbitrary free-text prose or operations unrelated to HarborCRM data, do not use this skill.
## Environment & API access
- HarborCRM is a **read-only** JSON API. Use `curl -s "<BASE><endpoint>" | python3 -m json.tool` (or pipe through `jq`).
- **Base URL:** use the remote base supplied in the environment access file (`<remote-env-url>`). Prompts may mention `127.0.0.1:8067` or a runner-supplied URL — prefer the remote base from the environment access file; never start or read a local `env/` service. If the runner explicitly hands you a different base URL at solve time, use that.
- Public GET endpoints (all return JSON; dataset is deterministic):
- `GET /health`
- `GET /api/events` · `GET /api/events/{event_id}` · `GET /api/events/{event_id}/orders` · `GET /api/events/{event_id}/badges` · `GET /api/events/{event_id}/sponsor_packages`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,953 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Output discipline (applies to every task)
- 1. Remote API usage
- 2. Policies (always fetch `/api/policies` and apply)
- 3. QUOTE tasks — two sub-families
- 4. RECONCILIATION tasks — won opportunity + milestones + revenue + events
- 5. Common misjudgments & exclusion rules (do NOT do these)
- 6. Worked decision flow (apply per task)
Files
SKILL.mdcore25,401 bytes
Read SKILL.md
# MedBridge Sales Ops — Account-Manager CRM Skill
Transferable operating procedure for two families of account-manager work against the
MedBridge Sales Ops remote API: **QUOTE decision packages** (reconcile a customer
RFQ/quote against catalog + freight + policies) and **RECONCILIATION packages**
(reconcile a won opportunity against invoices/payments/revenue-journals/events/vouchers).
The rules below are derived from the 5 train tasks and the live API records behind them.
Apply them to unseen tasks; do not memorize train answers.
---
## 0. Output discipline (applies to every task)
- Return **only valid JSON** matching `input/payloads/answer_template.json` exactly:When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,648 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API discipline
- 1. Task routing
- 2. QUOTE tasks
- 3. RECONCILIATION tasks
- 4. Controlled enum values (exact strings)
- 5. Owner-queue routing summary
- 6. Trust-API caveat — customer display name
- 7. Known gold-answer anomalies (do not over-fit)
- 8. Output discipline checklist
Files
SKILL.mdcore25,212 bytesSKILL.mdcore25,212 bytes
Read SKILL.md
# SKILL: MedBridge Sales Ops — Quote & Reconciliation Decision Packages
Executable field-craft for producing account/finance-ready JSON from the MedBridge Sales Ops remote API. Two task families share one environment; the output schema is dictated **per-task** by `input/payloads/answer_template.json`. Always read that template first and mirror its exact field names, nesting, and enum values — quote tasks come in at least three template variants and reconciliation tasks in at least two. Do NOT assume a single canonical schema.
## 0. Environment & API discipline
- Source of truth = the remote HTTP JSON API. Base URL is given to the runner as `API_BASE_URL` (env file: `<remote-env-url>`). Never start a local server; never read any `env/` directory.
- Endpoints: `GET /health`, `GET /api` (metadata), `GET /api/<collection>`, `GET /api/<collection>?<key>=<value>` (case-insensitive, AND-combined, 2dp numeric tolerance, matches nested keys), `GET /api/<collection>/<id>` **detail-by-id**, `GET /api/search?q=<text>` (cross-collection substring, ≤100 hits).
- **Detail-by-id is available ONLY for:** `customers`, `products`, `rfqs`, `quotes`, `freight-quotes`, `opportunities`. For `invoices`, `payments`, `revenue-journals`, `events`, `vouchers`, `policies` use the listing endpoint with a filter, or `/api/search`.
- Canonical filter keys: `freight-quotes?quote_id=…`; `invoices?opportunity_id=…`; `payments?opportunity_id=…`; `revenue-journals?opportunity_id=…`; `events?customer_id=…` (or `?opportunity_id=…`); `vouchers?customer_id=…`; `quotes?rfq_id=…`.
- Money = USD, 2 decimals. Dates = ISO `YYYY-MM-DD`. Use stable record IDs verbatim (e.g. `Q-TR-WC-1187`, `FR-WC-AIR`, `OPP-TR-HELIOS`, `CUST-MERIDIAN`).
- The shared business/"as-of" date across tasks is **2026-06-01** (the API `generated_at` and the quote_date used throughout). Use it as `as_of_date` when a reconciliation template requires one and the prompt does not state another.
- Reconcile the prompt's named entity (customer/quote/opportunity ID) against the API; **trust the API for IDs, amounts, dates, statuses, and tier facts**. One exception for the *display name* only — see §6.
- Output: only valid JSON matching the task's template. No markdown fences, no prose. Money as numbers (cent-level precision, e.g. `42480.0`). Nulls where the template allows (`null`), not omitted.
medbridge-sales-ops-reconcile
When your agent uses it
Reconcile MedBridge Sales Ops CRM tasks (B2B medical wholesale + implementation-services) against the remote Sales Ops API and emit the account-ready JSON answer. Covers QUOTE decision packages (EXW pricing, freight options, route risk, recommended mode, payment terms) and milestone RECONCILIATION packages (invoice/payment state, outstanding balance, revenue-recognition coverage, CRM follow-up routing, event/voucher linkage). Use whenever a task references MedBridge, quotes like Q-TR-*, RFQ-TR-*, opportunities like OPP-TR-*, milestones MS1/MS2/MS3, freight-quotes, revenue-journals, or the shared Sales Ops API base URL.
What it loads into context
Frontmatter
152 tokensAlways available to the agent
SKILL.md
6,856 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API access contract
- 2. Pick the task family and template variant
- 3. QUOTE family — common data fetch
- 4. RECONCILIATION family — common data fetch
- 5. Controlled enum value maps (consolidated)
- 6. Misjudgments & exclusions to avoid
- 7. Output discipline
Files
SKILL.mdcore24,900 bytes
Read SKILL.md
# MedBridge Sales Ops — Reconciliation & Quote Decision Skill
Executable rules for producing the single `answer.json` that matches `input/payloads/answer_template.json` exactly. Source of truth is the **remote Sales Ops HTTP API** (base URL supplied by the runner; treat `http://127.0.0.1:${PORT}` / `API_BASE_URL` placeholders in prompts as that same remote base). Trust API records over narrative wording for IDs and attributes, **except echo the customer name exactly as the task prompt states it** (see §7).
## 1. API access contract
- Base URL: the runner-provided `API_BASE_URL` (a remote host such as `<remote-env-url>`). Never start a local server; never read any `env/` directory.
- `GET /health` — sanity check.
- `GET /api` — lists `collections` and `endpoints`.
- `GET /api/<collection>` — list all records (`{collection, count, records}`).
- `GET /api/<collection>?<key>=<value>` — filter (case-insensitive, matches nested keys, numeric tolerance 2dp; multiple filters AND). Filter invoices/payments/revenue-journals/events/vouchers by `opportunity_id`; freight-quotes by `quote_id`.
- `GET /api/<collection>/<id>` — detail by id. **Detail-by-id works ONLY for:** `customers`, `products`, `rfqs`, `quotes`, `freight-quotes`, `opportunities`. For `invoices`, `payments`, `revenue-journals`, `events`, `vouchers`, `policies` use listing/filter (or `/api/search?q=...` which returns full records tagged with collection+id).
- `GET /api/search?q=<text>` — substring search across ALL collections (up to 100 hits). Use to resolve events/vouchers by code or id.
- Money = USD 2 decimals. Dates = ISO `YYYY-MM-DD`. Use stable record IDs verbatim.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,477 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API usage
- 1. QUOTE tasks — quote decision package
- 2. RECONCILIATION tasks — engagement/finance reconciliation
- 3. Cross-cutting pitfalls (from reflect feedback)
Files
SKILL.mdcore17,055 bytes
Read SKILL.md
# MedBridge Sales Ops — CRM Reconciliation & Quote Decision Skill
Transferable workflow rules for solving MedBridge Sales Ops CRM tasks against the
remote business API. Covers the two task families: **QUOTE** (reconcile a customer
RFQ/quote into an account-ready quote decision package) and **RECONCILIATION**
(reconcile a won opportunity into a finance-ready engagement reconciliation).
These are reusable SOPs and pitfalls — not answers. Always derive every value from
the live API records for the task's specific quote/RFQ/opportunity.
---
## 0. Environment & API usage
- Business records live ONLY behind the remote HTTP JSON API (`API_BASE_URL`). DoWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,684 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Overview
- API Usage Rules
- QUOTE Task Workflow
- QUOTE: Module RFQ Variant (indicative, EXW-only, no freight)
- RECONCILIATION Task Workflow
- Field Conventions
- Pitfalls
Files
SKILL.mdcore18,685 bytes
Read SKILL.md
# MedBridge Sales Ops CRM Skill
## Overview
This skill covers two families of account-ready tasks against the MedBridge Sales Ops API:
- **QUOTE tasks**: Reconcile a customer RFQ or quote revision against the API records and produce a quote decision package — EXW pricing, freight comparison with grand totals, route-risk flags, recommended transport mode, freight-validity warnings, and payment terms.
- **RECONCILIATION tasks**: Reconcile a won opportunity — CRM opportunity, milestone invoices, payments, revenue-recognition journals, events, and vouchers — and produce a finance-ready reconciliation with paid/unpaid state, outstanding balance, revenue-recognition coverage, accounting/collection/event-invitation follow-up actions, and contact linkage.
## API Usage Rules
The shared business environment is a **remote** HTTP JSON API (base URL from the task's `environment_access.md`).
### Detail-by-id endpoints (GET /api/<collection>/<id>)
Available **only** for: `customers`, `products`, `rfqs`, `quotes`, `freight-quotes`, `opportunities`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,555 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API Usage Rules
- 1. QUOTE Tasks — Workflow
- 2. RECONCILIATION Tasks — Workflow
- 3. Common Pitfalls Checklist
- 4. Output Discipline
Files
SKILL.mdcore17,079 bytesSKILL.mdcore17,079 bytesSKILL.mdcore17,079 bytesSKILL.mdcore17,079 bytes
Read SKILL.md
# MedBridge Sales Ops — CRM Reconciliation & Quote Decision Skill
Reusable workflow rules for producing account-ready JSON answers against the
MedBridge Sales Ops remote API. Covers the two task families: **QUOTE** tasks
(reconcile a customer RFQ/quote → quote decision package) and **RECONCILIATION**
tasks (reconcile a won opportunity → finance-ready reconciliation).
## 0. Environment & API Usage Rules
- The business source of truth is a **remote HTTP JSON API** at `API_BASE_URL`
(provided by the runner). Never read any local `env/` directory; never start a
local server. All business records come through `GET`/`POST` on the API.
- **Detail-by-id (`GET /api/<collection>/<id>`) is only available for:**
`customers`, `products`, `rfqs`, `quotes`, `freight-quotes`, `opportunities`.
For all other collections (`invoices`, `payments`, `revenue-journals`,When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,618 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API access
- 1. QUOTE TASK SOP
- 2. RECONCILIATION TASK SOP
- 3. Output field definitions & controlled enums (reference)
- 4. Common misjudgments & exclusions (checklist)
- 5. Step-by-step execution recipe
Files
SKILL.mdcore32,173 bytesSKILL.mdcore32,173 bytes
Read SKILL.md
# MedBridge Sales Ops — CRM Reconciliation & Quote Decision SOP
Transferable skill for two task families in the MedBridge Sales Ops domain:
- **QUOTE tasks** — reconcile a customer RFQ/quote against the API (customer, quote, product catalog tiers, freight quotes, policies) into an account-ready quote decision package (EXW pricing, freight options with grand totals, route risk flags, recommended transport mode, freight-validity warnings, payment terms).
- **RECONCILIATION tasks** — reconcile a won opportunity (CRM opportunity, milestone invoices, payments, revenue-recognition journals, events, vouchers) into a finance-ready reconciliation (paid/unpaid state, outstanding balance, revenue-recognition coverage, CRM follow-up tasks, event/voucher linkage).
All rules below were distilled from actually solving the 5 train tasks against the live remote API. They transfer to unseen test tasks.
---
## 0. Environment & API access
- Base URL: `<remote-env-url>` (from `environment_access.md`). All business records come ONLY through this HTTP API. Do not read any local `env/` directory.
- `GET /health` — health check. `GET /api` — lists `collections` and `endpoints`.
- `GET /api/<collection>` — list all records (`{collection, count, records}`).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,936 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API contract (memorize)
- 1. Universal workflow (every task)
- 2. QUOTE family — business rules
- 3. QUOTE family — TWO template variants (match exactly)
- 4. RECONCILIATION family — business rules
- 5. Common pitfalls / distractors (verified in train data)
- 6. Quick reference — controlled enum values
- 7. Worked train results (for self-check, NOT for copying — test tasks differ)
Files
SKILL.mdcore25,447 bytes
Read SKILL.md
# MedBridge Sales Ops — Quote & Reconciliation SOP
Transferable operating procedure for MedBridge Sales Ops CRM tasks (B2B medical wholesale + implementation-services). Two task families: **QUOTE** (reconcile a customer RFQ/quote against the API → account-ready quote decision package) and **RECONCILIATION** (reconcile a won opportunity's invoices/payments/revenue-journals/events → finance-ready reconciliation). All business records come ONLY from the remote HTTP API. Money is USD 2-decimals; dates are ISO `YYYY-MM-DD`; use stable record IDs exactly as they appear.
## 0. API contract (memorize)
- Base URL: from `environment_access.md` (`API_BASE_URL`). All GET.
- `GET /api` → metadata (lists collections + endpoints).
- `GET /api/<collection>` → `{collection, count, records[]}`._collections: `customers, products, rfqs, quotes, freight-quotes, policies, opportunities, invoices, payments, revenue-journals, events, vouchers`.
- `GET /api/<collection>?<key>=<value>` → filter (case-insensitive, matches nested keys, numeric tolerance 2dp, multiple filters AND together). Use this for `invoices?opportunity_id=...`, `payments?opportunity_id=...`, `freight-quotes?quote_id=...`, `revenue-journals?opportunity_id=...`.
- `GET /api/<collection>/<id>` → single record. **Detail-by-id works ONLY for:** `customers, products, rfqs, quotes, freight-quotes, opportunities`. For invoices/payments/revenue-journals/events/vouchers/policies use listing or filter endpoints (no detail-by-id).
- `GET /api/search?q=<text>` → substring search across ALL collections (≤100 hits, each tagged with `collection` + `id`). Great for finding an opportunity's full object graph by customer/opportunity name.
- **Pipeline `python3 -m json.tool`** to read. Prefer writing a fetch-script that dumps all needed endpoints to one file, then Read it (avoids dropped output).
### Critical API gotchasWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,539 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Environment & API usage rules
- 2. TASK FAMILY A — QUOTE (RFQ/quote → decision package)
- 3. TASK FAMILY B — RECONCILIATION (won opportunity → finance-ready)
- 4. Common misjudgments & exclusions checklist
- 5. Execution checklist (apply to every task)
Files
SKILL.mdcore28,162 bytesSKILL.mdcore28,162 bytes
Read SKILL.md
# MedBridge Sales Ops — Quote & Reconciliation SOP Skill
Transferable, executable standard-operating procedures for MedBridge Sales Ops CRM tasks. Distilled from hands-on reconciliation of the 5 official train tasks against the live remote API (no gold answers used). Applies to two task families: **QUOTE** (customer RFQ/quote → account-ready quote decision package) and **RECONCILIATION** (won opportunity → finance-ready engagement reconciliation).
The remote API is the single source of truth. Reconcile the prompt narrative against API records; **trust the API over narrative wording** when they conflict, but read the prompt carefully for *which* entity (quote id, opportunity id, customer id, product code) to reconcile.
---
## 1. Environment & API usage rules
Base URL is in `environment_access.md` (e.g. `<remote-env-url>`). All GET, all JSON. Pipe with `curl -s "..." | python3 -m json.tool`.
- `GET /health` → sanity (`{"ok": true, ...}`). `GET /api` → lists `collections` + `endpoints`.
- `GET /api/<collection>` → all records `{collection, count, records}`.
- `GET /api/<collection>?<key>=<value>` → filter (case-insensitive, matches nested keys, numeric tolerance 2dp). Multiple filters AND together.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,257 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API map
- 1. Diagnostic floors (inferrable thresholds)
- 2. Family A — Offline ticket batch resolution
- 3. Family B — Mobile contact-center triage
- 4. Family C — Enterprise export-complaint response package
- 5. Audit-math definitions (when test prompts request these)
- 6. Pitfalls and edge cases
- 7. Quick threshold summary (commit these numbers)
Files
SKILL.mdcore31,308 bytesSKILL.mdcore31,308 bytes
Read SKILL.md
# SCN_003 CRM Support-Operations Resolution Skill
Executable SOP for resolving offline-ticket batches, mobile contact-center queues,
and enterprise export-complaint response packages against the shared support console
API. Follow the procedures literally — record fields are the source of truth, never
assume values from the queue note or issue summary alone.
## 0. Environment & API map
Base URL (use this, not any 127.0.0.1 URL printed in a prompt):
`<remote-env-url>`
Health: `GET /health` → `{"ok": true}`
Catalog: `GET /api/catalog` (endpoint list + aggregate counts)
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,717 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API base
- 1. Family A — Ticket-batch resolution (train_001 / train_004)
- 2. Family B — Mobile triage (train_002 / train_005)
- 3. Family C — Enterprise export-complaint response package (train_003)
- 4. Pitfalls & conventions (all families)
Files
SKILL.mdcore24,288 bytesSKILL.mdcore24,288 bytes
Read SKILL.md
# SCN_003 CRM Support-Operations Skill
Reusable SOP for resolving CRM support-console batches across three task
families: (1) ticket-batch resolution, (2) mobile triage, (3) enterprise
export-complaint response packages. All facts below are distilled from gold
answers + live API probes and are intended for a downstream solver.
## 0. Environment & API base
- Prompts reference `http://127.0.0.1:8057` — ALWAYS replace with the real
base URL: **`<remote-env-url>`**.
- Health: `GET /health` -> `{"ok": true}`.
- Catalog: `GET /api/catalog` (endpoints + record counts only).
- Records are the source of truth; never assume a field exists. Probe to confirm.
scn003-crm-support-ops
When your agent uses it
SCN_003 CRM support-operations solver. Resolve ticket batches (fixed-line diagnostics + gates), mobile contact-center triage, and enterprise export-complaint response packages against the shared support-console API. Apply exact metric thresholds, gate order, root-cause->team map, mobile action decision tree, and enterprise package derivation rules.
What it loads into context
Frontmatter
81 tokensAlways available to the agent
SKILL.md
7,073 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API map
- 1. Family A — Ticket batch resolution (fixed-line)
- 2. Family B — Mobile contact-center triage
- 3. Family C — Enterprise export-complaint response package
- 4. Cross-family pitfalls
- 5. Output field definitions (quick reference)
Files
SKILL.mdcore27,318 bytes
Read SKILL.md
# SCN_003 CRM Support-Operations SOP
Executable procedure for three task families behind one support-console API. Always read `payloads/answer_template.json` first to detect which output schema is required, then follow the matching SOP. Return only JSON that conforms to the template. Preserve payload order (tickets) or ascending case_id order (mobile) exactly.
## 0. API map
Base URL: the harness-provided support console (training prompts show `http://127.0.0.1:8057`; the live base is `<remote-env-url>`). Use whatever base the harness supplies. All endpoints are GET, JSON.
Fixed-line / ticket domain:
- `GET /api/tickets/{ticket_id}` → `{ticket_id, account_id, service_area, service_type, subscribed_mbps, status, issue_summary, created_at}`
- `GET /api/accounts/{account_id}` → `{account_id, name, service_area, status, tier, authentication{last_login_status, last_login_at, account_recovery_status}}` (404 `{"error":"not_found"}` when account missing)
- `GET /api/diagnostics/{ticket_id}` → `{ticket_id, latency_ms, jitter_ms, bandwidth_mbps, root_causes[], started_at, completed_at}` (pre-troubleshooting metrics)
- `GET /api/troubleshooting/{ticket_id}` → `{ticket_id, post_latency_ms, post_jitter_ms, post_bandwidth_mbps, steps[], started_at, completed_at}` (post-troubleshooting metrics)
- `GET /api/outages` → list of `{outage_id, active, service_area, service_types[], eta_hours, impact_score, started_at}`reflect-3_attempt_01
When your agent uses it
SCN_003 CRM support-console analyst. Resolve same-day ticket batches, mobile triage queues, mobile-data recovery worklists, and enterprise export-complaint response packages by querying the shared support-console API and applying fixed diagnostic thresholds, gating order, root-cause routing, and per-family decision trees.
What it loads into context
Frontmatter
72 tokensAlways available to the agent
SKILL.md
4,826 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API Map
- 2. Diagnostic Thresholds (EXACT, converged values)
- 3. Root-Cause → Escalation Team Map
- 4. Family SOPs
- 5. Audit / Summary Formulas
- 6. Pitfalls
Files
SKILL.mdcore18,590 bytes
Read SKILL.md
# SCN_003 CRM Support Operations Analyst
Reusable rules for working a support-console workload across four task families. All evidence comes from the remote support-console API; never assume field values — fetch the records. Prompts may reference `127.0.0.1:8057`; always substitute the real base URL provided by the harness.
## 1. API Map
Base URL: use the harness-provided base URL (e.g. `<remote-env-url>`). Health: `GET /health`.
Public catalog: `GET /api/catalog` (endpoints + record counts only).
- `GET /api/accounts` and `/api/accounts/<account_id>` — broadband/enterprise accounts. Fields: `account_id`, `name`, `service_area`, `status` (`Active`/`Suspended`), `tier`, `authentication{last_login_status, account_recovery_status, last_login_at}`. A missing account returns `{"error":"not_found"}`.
- `GET /api/tickets` and `/api/tickets/<ticket_id>` — tickets. Fields: `ticket_id`, `account_id`, `service_type` (`internet`/`video`/`voice`), `service_area`, `subscribed_mbps`, `status`, `issue_summary`.
- `GET /api/outages?service_area=<area>` — outages for an area. Each: `outage_id`, `service_area`, `active` (bool), `service_types` (array), `eta_hours`, `impact_score`, `started_at`.
- `GET /api/diagnostics/<ticket_id>` — pre-troubleshoot snapshot. Fields: `bandwidth_mbps`, `latency_ms`, `jitter_ms`, `root_causes` (array), `started_at`/`completed_at`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,327 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API map
- 1. Ticket family SOP (offline-service ticket batch & queue review)
- 2. Mobile triage SOP (contact-center queue, full mobile)
- 3. Mobile-data recovery SOP (mobile-data worklist)
- 4. Enterprise package SOP (export-failure response)
- 5. Cross-family execution rules
Files
SKILL.mdcore20,344 bytesSKILL.mdcore20,344 bytes
Read SKILL.md
# SC003 CRM Support Operations Analyst
Resolve support-console operations across three task families using the shared support console API.
This is a train-only reflect skill: it contains reusable RULES only (no task answers, no judge hooks).
## 0. Environment & API map
Base URL (always use this; ignore any `127.0.0.1` URL in prompts):
`<remote-env-url>`
- `GET /health` -> `{"ok": true}`
- `GET /api/catalog` -> endpoint list + record counts (public).
- Accounts/tickets:
- `GET /api/accounts`, `GET /api/accounts/<account_id>`
- `GET /api/tickets`, `GET /api/tickets/<ticket_id>`reflect-3-crm-support-console
When your agent uses it
Resolve CRM support-console tasks (SCN_003) — ticket-batch classification, mobile-data triage, and enterprise export-complaint response packages — against a remote support-console API. Use when a task asks to classify service tickets / queue snapshots, triage mobile data cases, or build an enterprise export-failure response package and references a support console API.
What it loads into context
Frontmatter
85 tokensAlways available to the agent
SKILL.md
4,401 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API MAP (substitute real ids)
- FAMILY 1 — TICKET BATCH / QUEUE CLASSIFICATION
- FAMILY 2 — MOBILE TRIAGE / DATA RECOVERY
- FAMILY 3 — ENTERPRISE EXPORT-COMPLAINT RESPONSE PACKAGE
- PITFALLS
Files
SKILL.mdcore17,716 bytesSKILL.mdcore17,716 bytesSKILL.mdcore17,716 bytes
Read SKILL.md
# SCN_003 CRM Support Console — Resolution Skill
Reusable rules for three task families driven by a shared support-console API.
**Base URL**: use the base URL supplied by the harness/environment (the prompts mention a 127.0.0.1 address — ALWAYS substitute the provided base URL). Records are the source of truth; never assume field values — fetch them.
## API MAP (substitute real ids)
| Endpoint | Use |
|---|---|
| `GET /health` | liveness |
| `GET /api/catalog` | endpoint + record-count catalog (public) |
| `GET /api/accounts`, `GET /api/accounts/<account_id>` | account status, authentication, service_area, tier |
| `GET /api/tickets`, `GET /api/tickets/<ticket_id>` | ticket: account_id, service_area, service_type, subscribed_mbps |
| `GET /api/outages?service_area=<area>` | outages: active, service_area, service_types[], outage_id |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,966 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API map
- 1. Family 1 — Offline ticket batch resolution (train_001, train_004)
- 2. Family 2 — Mobile contact-center triage (train_002, train_005)
- 3. Family 3 — Enterprise export-complaint response package (train_003)
- 4. Audit math definitions (test tasks may request rich audit fields)
- 5. Pitfalls (do not trip on these)
- 6. Quick reference — derived constants
Files
SKILL.mdcore26,122 bytesSKILL.mdcore26,122 bytes
Read SKILL.md
# SCN_003 CRM Support-Console Resolution SOP
Executable SOP for the five CRM train families. All facts below were derived by
probing the live support-console API and cross-referencing records. Thresholds
marked **(inferred)** are committed numeric best-estimates reasoned from the
data; apply them as stated.
## 0. Environment & API map
- **Base URL:** `<remote-env-url>` — ALWAYS use this instead of any
`127.0.0.1:8057` / localhost URL mentioned in a task prompt.
- Health: `GET /health` -> `{"ok": true}`.
- Catalog: `GET /api/catalog` (endpoints + record counts only).
- Records are the source of truth. Never assume a field — fetch it.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,217 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API MAP
- 1. TICKET FAMILY — gating, diagnostics, thresholds, escalation
- 2. AUDIT MATH DEFINITIONS (ticket family)
- 3. MOBILE TRIAGE FAMILY — decision tree (train_002, train_005)
- 4. KEY DISCRIMINATORS & PITFALLS (mobile)
- 5. ENTERPRISE EXPORT-COMPLAINT FAMILY (train_003)
- 6. GENERAL EXECUTION RULES
Files
SKILL.mdcore30,719 bytes
Read SKILL.md
# SCN_003 CRM Support-Console Operations Skill
Executable SOP for resolving three families of support-console tasks against the shared API:
1. **Ticket batch resolution / queue quality** (offline service tickets) — train_001, train_004
2. **Mobile contact-center triage** (case queue / mobile-data worklist) — train_002, train_005
3. **Enterprise export-complaint response package** — train_003
All facts below were derived by probing the live API and reasoning through the train tasks. Numeric thresholds marked **[inferred]** are committed estimates; string conventions marked **[inferred]** are committed naming rules. Use them as-is.
---
## 0. API MAP
Base URL: the prompt says `http://127.0.0.1:8057` "unless the harness provides a different base URL". The harness-provided base for this environment is **`<remote-env-url>`**. Always use the harness-provided base if present; fall back to the prompt's literal base otherwise.
scn003-crm-support-console
When your agent uses it
Resolve SCN_003 CRM support-console tasks: ticket-batch resolution, mobile contact-center triage, and enterprise export-complaint response packages. Drive the remote console API, apply diagnostic floors + gating order, and emit the per-family JSON answer.
What it loads into context
Frontmatter
63 tokensAlways available to the agent
SKILL.md
8,366 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & base URL
- 1. API map
- 2. Ticket family — inferred diagnostic floors (CRITICAL)
- 3. Ticket family — gating order & resolution (train_001, train_004)
- 4. Mobile family — triage decision tree (train_002, train_005)
- 5. Enterprise family — export-complaint response package (train_003)
- 6. Audit-math definitions (ticket family — for audit-style output variants)
- 7. Output field definitions (schema recap)
- 8. Pitfalls
- 9. Threshold summary (inferred)
Files
SKILL.mdcore29,888 bytesSKILL.mdcore29,888 bytes
Read SKILL.md
# SCN_003 CRM Support Console — Executable SOP
Self-evolved skill. All thresholds marked **(inferred)** were derived by reasoning against the live API records (no gold answers). They are committed numeric estimates.
## 0. Environment & base URL
- **Base URL: `<remote-env-url>`** (ALWAYS use this; ignore any `http://127.0.0.1:8057` in prompts).
- Health: `GET /health` → `{"ok": true}`.
- Catalog: `GET /api/catalog` (endpoints + record counts).
- Records are the source of truth — never assume fields; fetch them.
## 1. API map
### Ticket / fixed-line domain (train_001, train_004)When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,372 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API access
- 1. Stable policy codes (emit EXACTLY these)
- 2. Precision rules (deterministic)
- 3. Data hygiene & sourcing rules
- 4. Retention risk signals (`RS-6`)
- 5. Risk score, levels, ranking, and actions
- 6. Net revenue exposure & segment summary (`EXP-6`)
- 7. Follow-up calendar (`CAL-5`)
- 8. Churn model conventions (`MOD-7`, `PRB-4`, `DEP-5`, `OUT-2`)
- 9. Per-archetype output schemas
- 10. Solver execution checklist
Files
SKILL.mdcore23,481 bytesSKILL.mdcore23,481 bytesSKILL.mdcore23,481 bytes
Read SKILL.md
# ApexCloud Retention Operations — Analytics Convention Skill
A reusable, executable convention layer for the ApexCloud Retention Operations API.
A test solver receives this SKILL + a task prompt + the environment base URL, and must
reproduce the conventions exactly. All rules below were verified against the live API and
the train answer keys. Apply them to whatever accounts / dates / quarters the test prompt
supplies.
---
## 0. Environment & API access
- **Base URL:** use the remote URL supplied to the solver (the `ENV_URL`). Ignore any
`http://127.0.0.1:...` placeholder inside a prompt body — the real service is the ENV_URL.
- All endpoints are read-only GET. Return JSON (`/exports/...` return raw CSV).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,105 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & endpoint usage
- 1. Global output conventions (deterministic precision)
- 2. Stable policy-code dictionary (emit exactly these values)
- 3. Data-source & hygiene rules
- 4. Retention risk model (RS-6) — reason codes
- 5. Primary action mapping (ACT-5) and next-touch calendar (CAL-5)
- 6. Task archetypes & output field definitions
- 7. Reason-code canonical order (emit in this relative order)
- 8. Quick checklist before returning JSON
Files
SKILL.mdcore21,932 bytesSKILL.mdcore21,932 bytesSKILL.mdcore21,932 bytesSKILL.mdcore21,932 bytesSKILL.mdcore21,932 bytes
Read SKILL.md
# ApexCloud Retention Operations — CRM Retention Analytics Skill
Executable experience for solving ApexCloud CRM retention-analytics tasks against the
ApexCloud Retention Operations API. A solver receiving only this file + a task prompt +
the environment URL must reproduce every convention below exactly.
## 0. Environment & endpoint usage
- The API base URL is given in the task's `ENV_URL.txt` (a remote HTTP endpoint). Task
prompts may print `http://127.0.0.1:8074` as an example — **ignore that literal** and use
the URL from `ENV_URL.txt`. Do not attempt to run any `setup.sh`.
- Public endpoints:
- `GET /api/health` — service/row-count sanity check.
- `GET /api/accounts` — list all accounts (44).
- `GET /api/accounts/<id>` — one account profile.apexcloud-retention-ops
When your agent uses it
Solve ApexCloud Retention Operations analytics tasks against the remote Retention Operations API. Encode and reproduce the exact business rules, data-hygiene conventions, policy codes, ranking/precision rules, churn-model spec, and per-archetype output schemas. Use for any CRM retention task (renewal risk queue / save plan, QBR metrics packet, receivables + pipeline operations digest, churn model validation + outreach ranking, high-touch retention action board / watchlist).
What it loads into context
Frontmatter
105 tokensAlways available to the agent
SKILL.md
10,124 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment
- 1. Universal precision & formatting rules
- 2. Universal data-hygiene conventions
- 3. Reason codes (definitions + fixed ordering)
- 4. Risk model, actions, exposure, calendar
- 5. Archetype field definitions & computation
- 6. Policy-code dictionary (EXACT values — emit these verbatim)
- 7. Worked gold reference (for self-check)
- 8. Execution checklist
Files
SKILL.mdcore35,504 bytesSKILL.mdcore35,504 bytes
Read SKILL.md
# ApexCloud Retention Operations — Solver Skill
This skill is the executable experience for the ApexCloud CRM retention-analytics evaluation. A future solver receives ONLY this file + a test prompt + the env URL and must reproduce every convention exactly. Read the test prompt, identify the archetype, apply the universal rules below, then the archetype-specific section.
## 0. Environment
- The API base URL is the REMOTE ApexCloud Retention Operations API. Read it from `ENV_URL.txt` in the task staging directory (a single line like `http://host:port`). Do NOT assume the `http://127.0.0.1:8074` written inside prompt bodies — that is a placeholder; the real base is in `ENV_URL.txt`.
- All endpoints are JSON over HTTP GET unless noted. Append paths to the base URL.
### Endpoints (all confirmed against the live schema)
| Endpoint | Returns | Key fields |
|---|---|---|
| `/api/health` | row_counts + status | sanity check; confirms dataset sizes |
| `/api/accounts` | `{accounts:[...]}` | `account_id`, `legal_name`, `account_aliases`, `segment`, `region`, `renewal_date`, `contract_tenure_months`, `crm_arr`, `billing_arr_current`, `lifecycle_status`, `product_plan` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,796 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. General principles
- 1. ARR / revenue source [VERIFIED train_001 + train_003]
- 2. Support-ticket hygiene [VERIFIED train_002]
- 3. NPS [VERIFIED train_003/data]
- 4. Overdue receivables [VERIFIED train_003 scored 1.0]
- 5. CRM matching [VERIFIED train_003 scored 1.0]
- 6. Pipeline / opportunities [VERIFIED train_003 scored 1.0]
- 7. HR summary [VERIFIED train_003 scored 1.0]
- 8. Event performance [VERIFIED train_003 scored 1.0]
- 9. Churn model [VERIFIED train_004 — model_validation block scored]
- 10. Risk-ranking & action boards [partially recoverable]
- 11. Retention board specifics (train_005 archetype) [VERIFIED train_005]
- 12. policy_codes & block-scored fields [NOT recoverable]
- 13. Archetype field definitions
- 14. What is recoverable vs not (calibration)
Files
SKILL.mdcore17,807 bytesSKILL.mdcore17,807 bytesSKILL.mdcore17,807 bytesSKILL.mdcore17,807 bytes
Read SKILL.md
# ApexCloud Retention Operations — Answer-Construction Skill
Reusable rules for building deterministic, judge-correct JSON answers against the
ApexCloud Retention Operations API. Derived from 3-round judge-feedback loops on
train_001..train_005.
Base URL and endpoints (data API, same host as judge):
`/api/health`, `/api/accounts`, `/api/accounts/<id>`,
`/api/accounts/<id>/metrics?start=YYYY-MM&end=YYYY-MM`,
`/api/accounts/<id>/tickets?start=YYYY-MM-DD&end=YYYY-MM-DD`,
`/api/accounts/<id>/nps?start=YYYY-MM-DD&end=YYYY-MM-DD`,
`/api/billing/snapshots`, `/api/finance/ar-aging?as_of=YYYY-MM-DD`,
`/api/opportunities?start&end®ion`, `/api/hr/summary?quarter®ion`,
`/api/events/performance?event&quarter`,
`/exports/churn/{train,validation,candidates}.csv`,reflect-3-attempt-02
When your agent uses it
ApexCloud Retention Operations — reusable rules and verified conventions for producing renewal-risk, QBR, receivables/pipeline, churn-model, and retention-board answers from the data API.
What it loads into context
Frontmatter
49 tokensAlways available to the agent
SKILL.md
4,679 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Environment & API
- Precision (deterministic, apply always)
- 1. ARR source **VERIFIED**
- 2. Support / ticket hygiene **VERIFIED**
- 3. NPS **VERIFIED**
- 4. Overdue receivables **VERIFIED**
- 5. CRM / A-R matching **VERIFIED**
- 6. Pipeline **VERIFIED**
- 7. HR & event ops context **VERIFIED**
- 8. Risk ranking & reason codes (risk + board archetypes) **BEST-GUESS**
- 9. primary_action & due dates **BEST-GUESS**
- 10. Churn model (validation + outreach ranking) — EXACT CONFIG **VERIFIED**
- 11. Net revenue exposure & segment summary (board archetype) **BEST-GUESS**
- 12. policy_codes — recoverable, rule-encoded **VERIFIED-RECOVERABLE**
- 13. Output field definitions by archetype
- 14. Common pitfalls
Files
SKILL.mdcore18,063 bytesSKILL.mdcore18,063 bytesSKILL.mdcore18,063 bytesSKILL.mdcore18,063 bytes
Read SKILL.md
# ApexCloud Retention Operations — Verified Conventions
Reusable rules for answering the five task archetypes served by the ApexCloud Retention Operations API. Every rule below was checked against the train judge; conventions marked **VERIFIED** produced correct deterministic fields (train_003 reached 1.0; train_002 deterministic block correct). Items marked **BEST-GUESS** could not be fully recovered from scalar feedback and should be applied with the stated logic.
## Environment & API
- Base URL: read from `ENV_URL.txt` (data API). Judge path from `JUDGE_PATH.txt`. All endpoints are read-only JSON.
- Endpoints: `/api/health`, `/api/accounts`, `/api/accounts/<id>`, `/api/accounts/<id>/metrics?start=YYYY-MM&end=YYYY-MM`, `/api/accounts/<id>/tickets`, `/api/accounts/<id>/nps`, `/api/billing/snapshots`, `/api/finance/ar-aging?as_of=YYYY-MM-DD`, `/api/opportunities`, `/api/hr/summary?quarter=YYYY-QN®ion=...`, `/api/events/performance?event=...&quarter=YYYY-QN`, `/exports/churn/{train,validation,candidates}.csv`, `/exports/account_metric_extract.csv`.
- `accounts` endpoint returns `{"accounts":[...]}`; `billing/snapshots` returns `{"snapshots":[...]}` (key `billing_arr`, field `posted: true`, quarterly `as_of` = 03-31/06-30/09-30/12-31); `ar-aging` returns `{"ar_aging":[...]}` (buckets `1_30, 31_60, 61_90, 90_plus`, keyed by `customer_name` = legal name, NO account_id field); opportunities have `account_id`, `stage`, `close_date`, `amount`, `state`.
## Precision (deterministic, apply always)
- Currency → 2 decimals. Percentages → 1 decimal. Counts and risk scores → integers. Churn probabilities → 3 decimals. NPS score → integer. Dates → `YYYY-MM-DD` / `YYYY-MM`.
## 1. ARR source **VERIFIED**
- `current_arr` = `billing_arr` from the **posted** billing snapshot whose **`as_of` == the assessment date**. Snapshots are quarterly; the assessment date (e.g. 2026-06-30) always matches a quarter-end `as_of`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,266 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Data API endpoints
- 1. UNIVERSAL CONVENTIONS (verified across train tasks)
- 2. ARCHETYPE-SPECIFIC RULES
- 3. POLICY CODES
- 4. OUTPUT FIELD REFERENCE (quick)
- 5. JUDGE-FEEDBACK READINGS (skill-generation notes)
Files
SKILL.mdcore19,695 bytesSKILL.mdcore19,695 bytesSKILL.mdcore19,695 bytes
Read SKILL.md
# ApexCloud Retention Analytics — Reflect-3 Skill
Reusable rules for solving ApexCloud CRM retention-analytics tasks against the
ApexCloud Retention Operations data API. Read the data API base URL from the
task prompt (the `http://127.0.0.1:8074` host in prompts is the documented
endpoint family; resolve to the live base URL given in the environment). These
rules are data/precision/labeling conventions — NOT gold answers and NOT a
test-time judge call.
## 0. Data API endpoints
- `/api/health` → row counts (use to sanity-check datasets: accounts=44,
account_metrics=528, billing_snapshots=176, ar_aging=196 (4 quarters × 49),
opportunities=114, hr_summary=16 (4 quarters × 4 regions),
event_performance=20, nps_responses=451, support_tickets=1595,
churn_train=180, churn_validation=60, churn_candidates=44).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,987 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment
- 1. Precision & formatting (apply to every task)
- 2. Data-hygiene rules (apply universally)
- 3. Retention-risk scoring & ranking (risk queue / board)
- 4. Action / reason-code dictionaries (controlled enums)
- 5. Per-archetype SOPs
- 6. Reason-code quick-reference by data test
- 7. Tenure risk direction (cross-task consistency)
- 8. Policy-code reference (inferred convention labels)
- 9. Common pitfalls (do not do these)
Files
SKILL.mdcore28,855 bytes
Read SKILL.md
# apex-retention-ops — ApexCloud Retention Operations SOP
Executable skill for CRM retention-analytics tasks against the ApexCloud
Retention Operations API. Covers five task archetypes: renewal risk queue,
QBR metrics packet, receivables + pipeline operations review, churn model
validation + outreach ranking, and the high-touch retention operations board.
All conventions below were reverse-engineered from the live API schemas and
the train answer templates (self-evolution: no gold answers were used). Where
an exact policy value could not be confirmed, the field name + governing rule
are stated so a solver applies the right policy.
---
## 0. EnvironmentWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,969 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Endpoint inventory & usage
- 1. Data hygiene rules (crisp, verifiable — apply always)
- 2. CRM matching rule (receivables → accounts)
- 3. Pipeline / opportunity rules
- 4. Risk scoring & ranking (renewal-risk queue + action board)
- 5. Churn model validation & outreach ranking (task-004)
- 6. QBR metrics packet (task-002)
- 7. Operations-review context (task-003 `ops_context`)
- 8. Precision & output conventions (apply to every archetype)
- 9. Policy codes (best inference — NO gold available)
- 10. Per-archetype output field definitions (from templates)
- 11. Reproducible SOP (apply per task)
Files
SKILL.mdcore21,781 bytesSKILL.mdcore21,781 bytes
Read SKILL.md
# ApexCloud Retention Operations — CRM Retention Analytics Skill
Self-evolved skill for the ApexCloud Retention Operations API. Covers the five
training archetypes: (1) renewal-risk queue, (2) QBR metrics packet,
(3) receivables + pipeline operations review, (4) churn-model validation +
outreach ranking, (5) high-touch retention action board. Conventions below are
inferred from live API data shapes and answer templates only — no gold answers.
Base URL is read from `ENV_URL.txt` (e.g. `<remote-env-url>`). Prompts
may reference `http://127.0.0.1:8074`; ALWAYS use the ENV_URL host, never the
127.0.0.1 placeholder.
---
## 0. Endpoint inventory & usageWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,639 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & base URL
- 1. Endpoint schema reference (verified)
- 2. Common parameters & date conventions
- 3. Data-hygiene & source-selection rules (load-bearing)
- 4. Pipeline (opportunities) rules
- 5. Renewal-risk scoring rubric (deterministic, integer 0-100)
- 6. Exposure formulas
- 7. Churn model recipe (fully reproducible)
- 8. Precision & output rules
- 9. Per-archetype output contracts (inferred from templates)
- 10. Policy-code conventions (inferred; no gold)
- 11. Execution checklist (apply per task)
- 12. Pitfalls & gotchas
Files
SKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytesSKILL.mdcore31,654 bytes
Read SKILL.md
# ApexCloud Retention Operations — CRM Retention Analytics Skill
Self-evolved, transferable SOP for the ApexCloud Retention Operations API. Encodes endpoint schemas, data-hygiene rules, ARR-source selection, pipeline semantics, a deterministic renewal-risk scoring rubric, the reproducible churn-model recipe, ranking/precision rules, per-archetype output contracts, and inferred policy-code conventions. No gold answers are stored; every rule is derived from the live API and answer-template shapes so it generalizes to held-out task variants.
## 0. Environment & base URL
- Base URL lives in `scratch/skill_generation/self_attempt_03/ENV_URL.txt`. Read it at run time; the host/port may change between evals. (The file is JSON-encoded text; parse the URL with `json.loads` or strip quotes/escapes, e.g. `<remote-env-url>`.)
- Every task hardcodes `http://127.0.0.1:8074` in its prompt as a default — IGNORE that literal; always use the URL from `ENV_URL.txt`. If the prompt gives a different assessment date / month set / account list, those task-specific values override the examples here.
- Call with `curl`. Most endpoints wrap results in an object: `{"accounts":[...]}`, `{"metrics":[...]}`, `{"tickets":[...],"count":N}`, `{"nps_responses":[...],"count":N}`, `{"snapshots":[...]}`, `{"ar_aging":[...]}`, `{"opportunities":[...]}`, `{"count":N,"hr_summary":[...]}`, `{"count":N,"event_performance":[...]}`. Always index the wrapper key, not the top-level list.
- `/api/health` returns per-resource row counts (useful to sanity-check export sizes): accounts 44, billing_snapshots 176, ar_aging 196, opportunities 114, support_tickets 1595, nps_responses 451, hr_summary 16, event_performance 20, churn_train 180, churn_validation 60, churn_candidates 44, account_metric_extract 528.
## 1. Endpoint schema reference (verified)
### /api/accounts and /api/accounts/<account_id>
`{"accounts":[{account_id, legal_name, display_name, account_aliases[], segment, region, product_plan, lifecycle_status, csm_owner, renewal_date (YYYY-MM-DD), contract_tenure_months (int), crm_arr (STALE — do not use), billing_arr_current (year-end snapshot value — do not use for assessment-date ARR)}]}`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,705 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API discovery and contract
- 2. Data shapes (field reference)
- 3. Cross-cutting rules (apply everywhere)
- 4. Task type A — Reimbursement-to-AP close (claim/bill/payment reconciliation)
- 5. Task type B — Vendor onboarding finance-risk release (KYC)
- 6. Task type C — Prepaid amortization close & GL reconciliation
- 7. Task type D — AP payment release after vendor account-change (release/hold/escalate)
- 8. Common misjudgments and exclusion rules
- 9. Controlled vocabularies (exact enum values required by answer templates)
- 10. Concrete solver SOP (recommended order for any task in this group)
Files
SKILL.mdcore30,345 bytesSKILL.mdcore30,345 bytesSKILL.mdcore30,345 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
Self-contained operating guide for solving finance expense-control tasks against the shared ERP
finance API. A solver needs only the API base URL (provided by the runner) plus this document.
## 1. API discovery and contract
### Base URL and discovery
- The runner supplies the API base URL (the local form `http://127.0.0.1:8005` and the remote form
`<remote-env-url>` serve the same data). Always treat the **remote API as the system of
record**; never rely on local files or stale snapshots as truth.
- Discover every available path with `GET {base}/endpoints`. It returns the path list and the
filtering contract.
- Health checks: `GET /health` and `GET /api/health`.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,516 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Contract
- Data Model Field Reference
- Task Type 1: Reimbursement-to-AP Close Review
- Task Type 2: Stale-Snapshot AP Reconciliation (Conference Reimbursement)
- Task Type 3: Vendor Onboarding KYC Release
- Task Type 4: Prepaid Amortization & GL Reconciliation
- Task Type 5: Payment Release after Account-Change Events
- Common Misjudgments / Exclusion Rules
- Payment-Priority Ranking (for AP payment release ordering)
- Query Order Cheat Sheet
Files
SKILL.mdcore20,814 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
## API Contract
Base URL: provided by the runner (e.g. `<remote-env-url>`). Discover with `GET /endpoints`.
### Endpoints (namespaced canonical paths)
| Path | Resource | Key filter params |
|------|----------|-------------------|
| `/api/claims` | Expense claims | `claim_id` |
| `/api/ap/bills` | AP bills (vendor invoices) | `claim_id`, `bill_id`, `vendor_id`, `status` |
| `/api/ap/payments` | AP payments | `bill_id`, `payment_id`, `vendor_id`, `status` |
| `/api/ap/aging` | AP aging summary | — |
| `/api/vendors` | Vendor master | `vendor_id` |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,844 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API access & discovery contract
- 2. Task archetypes & SOPs
- 3. Output field definitions & required enums
- 4. Compliance → hard_stop_flags mapping (Archetype C)
- 5. Computation rules that are easy to get wrong
- 6. Common misjudgments & exclusion rules
- 7. Recommended query order (general)
Files
SKILL.mdcore26,515 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
A self-contained playbook for solving ERP finance expense-control tasks against the shared
remote API. The solver only needs the remote API base URL (provided per task in
`environment_access.md` or the prompt). Discover everything else from the API itself.
> NEVER call any `/api/judge` endpoint. NEVER read local `env/` sources, evaluator files, or
> test/train answer keys. The remote API is the only system of record. Staged local payloads
> (answer templates, batch/scope files, stale CSVs) are inputs only.
---
## 1. API access & discovery contract
- **Base URL:** the runner-provided remote URL (e.g. `<remote-env-url>`). Prompts mayWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,179 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API contract
- 2. General reconciliation SOP
- 3. Expense-claim reimbursement close (claim -> AP bill -> payment)
- 4. Vendor onboarding finance-risk release (per business)
- 5. Prepaid close reconciliation (per account, per period)
- 6. AP batch reconciliation from a stale export (claim, bill, payment, close-log)
- 7. Payment release after vendor account-change (per business)
- 8. Cross-task pitfalls (sharpened by judge feedback)
Files
SKILL.mdcore13,320 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
Reusable workflow rules for expense-control / AP / prepaid-close / vendor-compliance
tasks against the shared ERP finance API. Distilled from reflect training.
## 1. API contract
- Base URL: `<remote-env-url>` (runner-provided; treat remote API as system of record).
- `GET /endpoints` lists all paths. All list endpoints support exact-match query params by
field name, plus `limit`/`offset`.
- Resource paths (use the `/api/...` form):
- Claims: `GET /api/claims?claim_id=<id>`
- AP bills: `GET /api/ap/bills?claim_id=<id>` (also queryable by bill_id)
- AP payments: `GET /api/ap/payments?bill_id=<bill_id>`
- Vendors: `GET /api/vendors?vendor_id=<id>`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,043 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Ground rules
- 1. API contract
- 2. Claims / AP / payment reconciliation (reimbursement-to-AP close)
- 3. Stale-AP-export batch review (conference reimbursement)
- 4. Vendor onboarding / compliance release (finance-risk)
- 5. Payment-release risk review (post account-change)
- 6. Prepaid close (straight-line amortization reconciliation)
- 7. Common misjudgments to avoid
Files
SKILL.mdcore12,335 bytesSKILL.mdcore12,335 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
Reusable workflow rules for solving ERP finance expense-control tasks against the shared remote ERP API. Distilled from reflect-loop practice on train tasks. These are workflow rules, not per-task answers.
## 0. Ground rules
- Use ONLY the remote ERP Web API (system of record). Do NOT read local `env/` source files or test fixtures.
- Report all currency in USD to 2 decimals unless a task explicitly says otherwise. Follow the `answer_template.json` `unit`/`precision` fields (they override ambiguous prose). Example: a template that says `unit: USD, precision: 2` means dollars with 2 decimals — do NOT convert to cents even if prose says "cents".
- Sort every ID list exactly as the template specifies (usually ascending by ID; digits sort before uppercase letters, so `CLM-2025-00xx` < `CLM-2025-FIN-xx` < `CLM-2025-OPS-xx`; `PPD-2025-*` < `PPD-AUR-*`).
- Preserve list orders that the template says must match an input payload order (e.g. `selected_invoice_ids` "same order as scope file").
- Query specific records by exact-match params (claim_id, bill_id, business_id, account, period, prepaid_invoice_id, vendor_id). Do not pull whole collections.
## 1. API contract
- Base URL + `GET /endpoints` lists all paths and the filtering contract.
- Health: `GET /health`, `GET /api/health`.
- Resources (namespaced `/api/...` preferred; bare aliases also exist): `/api/claims`, `/api/ap/bills`, `/api/ap/payments`, `/api/vendors`, `/api/close/logs`, `/api/prepaids/invoices`, `/api/prepaids/gl-balances`, `/api/compliance/objects`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
2,535 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. ERP API contract
- 2. Claim / AP close-batch reconciliation (reimbursement-to-AP)
- 3. Compliance / vendor onboarding & payment-release review
- 4. Prepaid close reconciliation
- 5. Common misjudgments (avoid)
- 6. Output hygiene
Files
SKILL.mdcore10,405 bytes
Read SKILL.md
# ERP Finance Expense-Control Skill (task_group_005)
Reusable workflow rules for solving ERP finance expense-control / close / release-control tasks against the shared ERP API. Distilled from reflect-loop practice on train tasks. Contains no test answers and no test-time judge usage.
## 1. ERP API contract
- Remote API base URL: `<remote-env-url>` (the system of record; do not read local `env/` files).
- Discover paths with `GET /endpoints`. Health: `GET /health`, `GET /api/health`.
- Filtering: exact-match query parameters by field name, plus `limit`/`offset` for pagination. Always query by the specific ID/param (e.g. `?claim_id=CLM-...`, `?business_id=BUS-...`, `?bill_id=AP-...`, `?prepaid_invoice_id=PPD-...`, `?account=1250&period=2025-03`), never whole collections.
- Namespaced resource paths:
- Claims: `/api/claims`
- AP: `/api/ap/bills`, `/api/ap/payments`, `/api/ap/aging`
- Vendors: `/api/vendors`
- Compliance: `/api/compliance/objects` (flat object with all fields) + detail slices `/api/compliance/{profile,ownership,registry,screening,bank,risk}/{business_id}`
- Prepaids: `/api/prepaids/invoices`, `/api/prepaids/gl-balances`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,480 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API contract
- 2. Reimbursement-to-AP close (claim/bill/payment reconciliation)
- 3. Vendor onboarding / KYC release (compliance-derived decisions)
- 4. Payment release after account-change events
- 5. Prepaid amortization & GL reconciliation
- 6. Currency, ordering & format conventions
- 7. Common misjudgments / exclusion rules (watch-list)
- 8. SOP — query order per task type
Files
SKILL.mdcore21,523 bytes
Read SKILL.md
# SKILL.md — task_group_005 ERP Finance Expense-Control (SELF mode)
Executable, transferable rules for the task_group_005 ERP finance/compliance API.
Derived by working the 5 train tasks from the remote API directly (no gold answers).
This file contains RULES and FIELD MAPPINGS, not per-task answer cheat-sheets.
---
## 1. API contract
- Base URL is given in `environment_access.md` (a single host:port). The local `http://127.0.0.1:8005`
mentioned in some prompts is an alias of the same system; ALWAYS treat the runner-provided
base URL / `environment_access.md` value as system of record. Do NOT read any local `env/` source.
- Discovery: `GET /endpoints` returns the authoritative path list. Two alias trees exist —
`/api/<resource>` and `/<resource>` — both return identical data. Prefer the `/api/...` forms.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,426 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API Contract
- 2. Record Schemas & Controlled Vocabularies
- 3. Business Rules & SOPs by Task Type
- 4. Common Misjudgments & Exclusion Rules
- 5. Quick Reference — Query Order Cheat-Sheet
Files
SKILL.mdcore24,939 bytes
Read SKILL.md
# SKILL.md — ERP Finance Expense-Control (task_group_005)
Transferable executable skill for solving ERP finance expense-control tasks against the shared remote API. Distilled from independent analysis of five train tasks (reimbursement-to-AP close, vendor intake/KYC, prepaid amortization & GL reconciliation, stale-snapshot AP payment board, account-change payment release). No gold answers or judge feedback were used.
---
## 1. API Contract
**Base URL:** `<remote-env-url>` (use this, not 127.0.0.1 which may appear in prompts).
**Discovery:** `GET /endpoints` returns all paths + the filtering contract.
**Namespaced resource paths (use the `/api/...` form):**
| Resource | Path | Alt path | ~Total |When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,231 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. API contract
- 2. Data model — fields that matter
- 3. Task A — Reimbursement-to-AP close review
- 4. Task B — Vendor onboarding / KYC release board
- 5. Task C — Prepaid amortization & GL reconciliation close
- 6. Task D — Payment-release risk review after account-change events
- 7. Common misjudgments & exclusion rules (apply across all task types)
- 8. Output field & controlled-vocabulary reference
- 9. Currency / precision conventions & the "USD cents" trap
- 10. Concrete SOPs (quick reference)
Files
SKILL.mdcore32,148 bytesSKILL.mdcore32,148 bytesSKILL.mdcore32,148 bytes
Read SKILL.md
# SKILL — task_group_005 ERP Finance Expense-Control Close
Transferable, executable experience for solving ERP finance expense-control tasks in the
`task_group_005` shared API environment. Covers four task archetypes that recur across
train/test pairs:
- **A. Reimbursement-to-AP close review** (claim → AP bill → payment reconciliation; payable/blocked/paid partition; stale AP-snapshot correction).
- **B. Vendor onboarding / KYC release board** (compliance hard-stop flags, UBO reporting, release decisions).
- **C. Prepaid amortization & GL reconciliation close** (straight-line monthly schedule vs. GL ending balance, variance + data-quality flags).
- **D. Payment-release risk review after vendor account-change events** (compliance + vendor + bank/tax/license/risk gating, release/hold/escalate).
The rules below are derived from working the 5 train tasks against the live API. They are
**rules and decision criteria**, not a lookup of any answer set.
---When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
9,178 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API reference
- 1. Cross-cutting data model & linkage
- 2. Reconciliation primitives (used by every family)
- 3. Family A — Sourcing nomination readiness packet
- 4. Family B — Receiving-control closeout
- 5. Family C — AP close / vendor-balance + hold/release settlement
- 6. Family D — Change-control contract amendment
- 7. Cross-family pitfalls & common misjudgments
- 8. Final output discipline
Files
SKILL.mdcore36,024 bytesSKILL.mdcore36,024 bytesSKILL.mdcore36,024 bytes
Read SKILL.md
# SKILL — ProcureOps Procurement Supplier & Receiving Control (task_group_006)
Transferable, executable experience for solving ProcureOps procurement-control tasks against the
shared ProcureOps API. Covers four task families that appear in this scenario:
- **Family A — Sourcing nomination readiness packet** (program scope + as-of date, budget/readiness, per-SKU supplier commercial decisions, evidence/blocker sets, committee action).
- **Family B — Receiving-control closeout** (batch identity, receipt scope, ordered↔received↔billed reconciliation, contract/PO/invoice price match, AP hold/release, financial totals, supplier-risk overlay, **chargeback release file**).
- **Family C — AP close / vendor-balance + hold/release settlement** (invoice-level hold/release, qty-variance %, totals & scheduled payments & net balance, vendor balance rows, program close totals, payment hold/release queues).
- **Family D — Change-control contract amendment** (contract status/price/qty, usage & headroom/ceiling, program-budget incremental exposure, requisition approval state, supplier risk, hold actions, final decision).
This is experience, not a restatement of any answer. Follow the recipes; verify every value against the API.
---
## 0. Environment & API referenceWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,207 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API conventions (read first)
- 1. Query recipes (by goal)
- 2. Cross-cutting reconciliation rules
- FAMILY A — Sourcing nomination readiness (train_001 → test_003)
- FAMILY B — Receiving-control closeout (train_002 → test_001 / test_004)
- FAMILY C — AP close / vendor-balance + hold-release (train_003 → test_002)
- FAMILY D — Change-control contract amendment (train_004 → test_005)
- 3. Cross-cutting common misjudgments (avoid)
- 4. Output compliance checklist
Files
SKILL.mdcore24,524 bytesSKILL.mdcore24,524 bytesSKILL.mdcore24,524 bytes
Read SKILL.md
# SKILL — ProcureOps ERP Procurement Reconciliation (task_group_006)
Transferable execution skill for the ProcureOps procurement benchmark. Covers four task families: sourcing nomination readiness, receiving-control closeout, AP close / vendor-balance + hold-release, and change-control contract amendment. All rules below were verified against the live API on 2026-07-02.
## 0. API conventions (read first)
- Base URL: `<remote-env-url>` (prompts say `127.0.0.1:8006` — same service at the remote host).
- List: `GET /<coll>` → `{"count":N,"results":[...]}`. **No `_limit`/pagination param** — unknown params are treated as field filters; an unknown param name returns 0 results. Only filter by real field names.
- By id: `GET /<coll>/<id>` → the record object.
- Field filters match **substring, case-insensitive**, including nested list values (e.g. `?po_id=PO-AX17-4481`). Combine `?field=value` with `&start=YYYY-MM-DD&end=YYYY-MM-DD` for the collection's date field (inclusive).
- Each collection has ONE date field that `start`/`end` apply to: receipts→`receipt_date`, ap_invoices→`invoice_date`, payments→`scheduled_date`, vendor_risk_events→`event_date`, purchase_orders→`order_date`, budget_snapshots→`snapshot_date`, approval_events→`event_date`.
- Do NOT pull whole collections. Fetch by id, or by one foreign-key filter (`po_id`, `contract_id`, `supplier_id`, `invoice_id`, `object_id`, `program_id`, `sku`), then filter client-side.
### Reconciliation chain (memorize)
PO → receipts (same `po_id`, multiple allowed) → invoices (`invoice.po_id` AND/OR `invoice.receipt_id`) → payments (`payment.invoice_id`).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,544 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API mechanics
- A. Sourcing nomination readiness (train_001 pattern; test_003)
- B. Receiving-control closeout (train_002 pattern; test_001/test_004)
- C. AP close / vendor balance + hold-release (train_003 pattern; test_002)
- D. Change-control contract amendment (train_004 pattern; test_005)
- Universal pitfalls & misjudgments
Files
SKILL.mdcore29,792 bytesSKILL.mdcore29,792 bytes
Read SKILL.md
# ProcureOps ERP Procurement Control — Solver Skill (task_group_006)
Executable experience for the ProcureOps shared API. Distilled from 5 train tasks
across 4 families: A sourcing-nomination readiness, B receiving-control closeout
(+ chargeback-release sub-pattern), C AP close / vendor balance + hold-release,
D change-control contract amendment. Recipe-oriented; not a restatement of train answers.
## 0. Environment & API mechanics
- Base URL: `<remote-env-url>` (prompts' `http://127.0.0.1:8006` == this service). Use `curl`.
- By id: `GET /<coll>/<id>` -> the record object. List: `GET /<coll>` -> `{"count":N,"results":[...]}`.
- There is NO pagination / `_limit` param. Unknown query params are treated as **field filters**
(substring, case-insensitive, including nested list values) and a non-matching filter returns 0.
Do not invent params like `limit`/`page`. Use real field filters: `?po_id=`, `?supplier_id=`,
`?program_id=`, `?invoice_id=`, `?object_id=`. Date ranges: `?start=YYYY-MM-DD&end=YYYY-MM-DD`When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,367 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API recipe (all families)
- Family A — Sourcing nomination readiness (program packet, as-of date)
- Family B — Receiving-control closeout (single batch)
- Family C — AP close / vendor-balance + hold/release (invoice slice)
- Family D — Change-control contract amendment
- Family B-variant — Receiving/AP release+hold file (multiple POs/receipts/invoices + chargeback register)
- Cross-family common mistakes to avoid
Files
SKILL.mdcore26,769 bytesSKILL.mdcore26,769 bytesSKILL.mdcore26,769 bytes
Read SKILL.md
# ProcureOps ERP Procurement Solver Skill (task_group_006)
Reusable workflow rules for solving ProcureOps procurement-control tasks against the
shared API. The API is the source of truth; local memos only name anchors (PO/receipt/
invoice/contract IDs, dates, as-of dates, business-control rates). All amounts are USD
unless a template field states otherwise; round numbers to the precision each field demands.
## 0. Environment & API recipe (all families)
- Remote API base: `<remote-env-url>` (task prompts say `127.0.0.1:8006` — same service).
- `GET /health`, `GET /manifest` (record counts + anchor ids, no answer keys).
- By id: `GET /<coll>/<id>` -> the record object. List: `GET /<coll>?<field>=<val>` ->
`{"count":N,"results":[...]}`. Field filters are substring/case-insensitive (incl. nested
list values); blanks ignored. `start`/`end` filter the collection date field (inclusive).
There is NO pagination param; unknown query params become field filters returning 0.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,227 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API mechanics
- Universal solver workflow
- Family A — Sourcing nomination readiness (scope + as-of, budget, per-SKU supplier decision, evidence/blockers, committee action)
- Family B — Receiving-control closeout (batch identity, receipt scope, ordered/received/billed reconciliation, price match, AP hold/release, financials, supplier-risk overlay, chargeback)
- Family C — AP close / vendor-balance + hold/release (invoice hold/release, qty-variance %, totals & scheduled payments & net balance, vendor balance rows, program close totals, payment queues)
- Family D — Change-control contract amendment (contract status/price/qty, usage & headroom/ceiling, program-budget incremental exposure, requisition approval state, supplier risk, hold actions, final decision)
- Cross-family — Receiving/AP release file (per-invoice release decision net of chargebacks, receiving exceptions, chargeback netting, source classification, followups)
- Quick decision-logic reference (confirmed by judge feedback)
Files
SKILL.mdcore25,967 bytesSKILL.mdcore25,967 bytesSKILL.mdcore25,967 bytes
Read SKILL.md
# ProcureOps Procurement Control — Solver Skill (task_group_006)
Reusable workflow rules for solving Procurement Supplier & Receiving Control tasks against the shared ProcureOps API. Recipe-oriented experience; not candidate answers. Applies to all four task families: A Sourcing nomination, B Receiving-control closeout, C AP close / vendor-balance, D Change-control amendment, plus the cross-family Receiving/AP release file.
## 0. Environment & API mechanics
- Remote API (system of record): `<remote-env-url>`. Prompts that say `http://127.0.0.1:8006` mean THIS SAME remote service — always call the remote URL.
- Collections (GET list -> `{"count":N,"results":[...]}`; GET `/<coll>/<id>` -> the record object): `programs`, `suppliers`, `items`, `contracts`, `purchase_requisitions` (aliases `/purchase-requests`, `/purchase-requisitions`), `purchase_orders` (alias `/purchase-orders`), `receipts`, `ap_invoices` (alias `/ap/invoices`), `payments` (alias `/ap/payments`), `approval_events` (alias `/approvals`), `budget_snapshots` (alias `/budgets`), `vendor_risk_events` (alias `/vendor-risks`).
- Query filters match substring/case-insensitive (incl. nested list values); blank values ignored. `start`/`end` filter the collection's date field inclusively. There is NO pagination/`_limit` param — unknown params become field filters returning 0. Fetch by id from the memo/template; do not dump whole collections.
- `GET /manifest` gives record counts + anchor ids (no task ids / answer keys). `GET /health` confirms liveness.
- Reconciliation chain: Program -> contracts/requisitions/POs/budget_snapshots/invoices(via po.program_id). PO -> receipts(same po_id, multiples allowed) -> invoices(invoice.po_id / invoice.receipt_id) -> payments(payment.invoice_id). Contract(contract_id) <-> PO(po.contract_id) <-> PO line unit_price <-> invoice line unit_price (contract.unit_price is the price-match anchor).
- Units: USD amounts in records are DOLLARS. Read each answer_template's per-field unit/precision (some want cents, most want 2-decimal USD; qty 2 decimals; pct 1 decimal; ratios 4 decimals). Sort every ID list ascending; list fields marked "set" are compared as sets.
## Universal solver workflow
1. Read the prompt + the task-local memo/packet + `answer_template.json`. Extract: program_id, as-of/review date, target ids (POs/receipts/invoices/requisition/contract/supplier/sku), and any business-control rules (tax rate, approval good-actions, ceiling/budget exposure definitions, cancel-exclusion, supplier-watch policy).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,783 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API mechanics (all families)
- A. Sourcing nomination readiness (train_001 / test_003)
- B. Receiving-control closeout (train_002 / test_001 / test_004)
- B-variant. Receiving + AP release with chargebacks (train_005)
- C. AP close / vendor balance + hold/release (train_003 / test_002)
- D. Change-control contract amendment (train_004 / test_005)
- Cross-family pitfalls (judge-taught)
Files
SKILL.mdcore22,521 bytesSKILL.mdcore22,521 bytes
Read SKILL.md
# ProcureOps ERP Procurement Solver Skill (task_group_006)
Reusable workflow rules for solver agents on unseen test tasks in this procurement
benchmark. Recipe-oriented; NOT candidate answers. The remote API
`<remote-env-url>` is the source of truth (task prompts' `127.0.0.1:8006` =
this same service). Do NOT call any judge endpoint at test time.
## 0. API mechanics (all families)
- `GET /<collection>` -> `{"count":N,"results":[...]}`. `GET /<collection>/<id>` -> record object.
- Collections: `programs`, `suppliers`, `items`, `contracts`, `purchase_requisitions`
(alias `/purchase-requests`), `purchase_orders` (alias `/purchase-orders`),
`receipts`, `ap_invoices` (alias `/ap/invoices`), `payments` (alias `/ap/payments`),
`approval_events` (alias `/approvals`), `budget_snapshots` (alias `/budgets`),
`vendor_risk_events` (alias `/vendor-risks`).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,035 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API mechanics
- 1. Cross-cutting reconciliation rules
- 2. Family A — Sourcing nomination readiness (train_001 / test_003)
- 3. Family B — Receiving-control closeout + AP release/chargeback (train_002, train_005 / test_001, test_004)
- 4. Family C — AP close / vendor balance / hold-release (train_003 / test_002)
- 5. Family D — Change-control contract amendment (train_004 / test_005)
- 6. Cross-family misjudgment checklist
Files
SKILL.mdcore31,756 bytesSKILL.mdcore31,756 bytesSKILL.mdcore31,756 bytes
Read SKILL.md
# SKILL — ProcureOps ERP Procurement Reconciliation (task_group_006)
Executable recipes for the four task families in this group. The API is the source of truth; the local memo/packet only names IDs and business rules. Always re-derive state from the live API, never trust a memo number.
## 0. Environment & API mechanics
- Base URL: `<remote-env-url>`. Prompts that say `http://127.0.0.1:8006` mean THIS SAME service — substitute the remote URL.
- `GET /<collection>` → `{"count":N,"results":[...]}`. **No pagination, no `_limit`.** Unknown query params become substring/case-insensitive field filters (they also match inside nested list values), so an unknown filter returns 0 — only filter on real field names.
- `GET /<collection>/<id>` → the record object.
- Useful link filters: `?po_id=`, `?supplier_id=`, `?program_id=`, `?contract_id=`, `?sku=`, `?invoice_id=`, `?receipt_id=`, `?object_id=` (for approval_events).
- `start`/`end` filter the collection's *primary date field* inclusively. **Do not rely on this for as-of scoping** — instead fetch by link filter (e.g. `?po_id=X`) then filter client-side by the relevant date `<= as_of`. This avoids ambiguity about which date field is filtered.
- Collections: `programs, suppliers, items, contracts, purchase_requisitions, purchase_orders, receipts, ap_invoices, payments, approval_events, budget_snapshots, vendor_risk_events`.
- Amounts in API records are USD **dollars** (floats). Most answer templates want USD **cents** for unit prices OR dollars-rounded-to-cents for totals — read each template field's stated unit. Quantities carry 2 decimals.
## 1. Cross-cutting reconciliation rulesWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
9,443 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. API access rules (read first — these bite)
- 1. Cross-cutting reconciliation & conventions
- 2. Family A — Sourcing nomination readiness (train_001 / test_003)
- 3. Family B — Receiving-control closeout (train_002 / test_001 / test_004)
- 4. Family C — AP close / vendor-balance + hold-release (train_003 / test_002)
- 5. Family D — Change-control contract amendment (train_004 / test_005)
- 6. Cross-family: Family B+D hybrid — Receiving/AP release & chargeback (train_005 / test-style)
- 7. Quick reference — common misjudgments to avoid
- 8. Worked ID map (anchors, for orientation — NOT answers)
Files
SKILL.mdcore37,814 bytesSKILL.mdcore37,814 bytesSKILL.mdcore37,814 bytes
Read SKILL.md
# SKILL — ProcureOps Procurement Reconciliation (task_group_006)
Executable recipes for the four task families in this benchmark. The remote API is
the single source of truth; local memos only name the target IDs and business rules.
Work each task by: (1) read memo → extract target IDs + business controls + as-of date,
(2) fetch each linked API record by exact id, (3) reconcile, (4) fill the template.
## 0. API access rules (read first — these bite)
- Base URL: `<remote-env-url>`. Task prompts say `127.0.0.1:8006` = same service.
- List: `GET /<coll>` → `{"count":N,"results":[...]}`. By id: `GET /<coll>/<id>` → the record object.
- **FILTERS REQUIRE THE FULL EXACT ID STRING.** `?po_id=PO-AX17-4481` works; `?po_id=AX17`
and `?po_id=4481` return `count:0`. Do NOT trust "substring" claims — always pass the
complete id. Unknown query params are silently treated as field filters and yield 0.
- Date range on a collection's date field: `?start=YYYY-MM-DD&end=YYYY-MM-DD` (inclusive).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,170 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API (verified 2026-07-02)
- Golden rules (read every time)
- Family A — Sourcing nomination readiness (train_001 / test_003)
- Family B — Receiving-control closeout
- Family C — AP close / vendor-balance + hold/release (train_003 / test_002)
- Family D — Change-control contract amendment (train_004 / test_005)
- Cross-cutting exclusion & misjudgment rules (all families)
- Quick ID anchor参考 (do NOT hardcode; verify live)
Files
SKILL.mdcore28,362 bytesSKILL.mdcore28,362 bytesSKILL.mdcore28,362 bytesSKILL.mdcore28,362 bytes
Read SKILL.md
# SKILL — Procurement Supplier & Receiving Control (task_group_006)
Executable recipes for solver agents on unseen test tasks in this benchmark. The
benchmark has FOUR task families (A sourcing-nomination, B receiving-control
closeout, C AP close / vendor-balance, D change-control amendment). Test tasks
map to these families; train_001/003↔test_003, train_002↔test_001/test_004,
train_003↔test_002, train_004↔test_005 (families overlap). Apply the matching
family section plus the cross-cutting rules at the bottom.
## 0. Environment & API (verified 2026-07-02)
- Remote API (source of truth): `<remote-env-url>`. Prompts' `127.0.0.1:8006`
== this same service. Use curl.
- List: `GET /<coll>` -> `{"count":N,"results":[...]}`. By id: `GET /<coll>/<id>` ->
the object directly. Field filters match substring/case-insensitive, includingWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,820 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Meta-SOP (do this first, every task)
- 1. The Remote ERP API
- 2. Task families and their output shapes
- 3. Core business rule: effective (available) stock
- 4. Decision precedence (applies across families)
- 5. Stock classification & line outcomes
- 6. Shipping quote (Families A, and any task that requests quotes)
- 7. Family B — BOM replenishment algorithm
- 8. Family C — Supplier incident scorecard
- 9. Family E — Quality-hold / replenishment-control review
- 10. Common misjudgments — explicit exclusions
- 11. Reusable checklist before submitting
Files
SKILL.mdcore26,589 bytesSKILL.mdcore26,589 bytesSKILL.mdcore26,589 bytesSKILL.mdcore26,589 bytesSKILL.mdcore26,589 bytesSKILL.mdcore26,589 bytes
Read SKILL.md
# Northwind Components ERP Fulfillment — Solver Skill
Transferable operating procedure for the Northwind Components ERP fulfillment
evaluation. Distilled from the 5 train tasks and verified against the live API.
Use this at TEST time to solve unseen variants of the same task families.
This skill is self-contained. It never instructs calling any judge/scoring
endpoint — only the public ERP API.
---
## 0. Meta-SOP (do this first, every task)
1. **Read the prompt + every payload file (`input/payloads/*`) and
`answer_template.json` literally.** Memos frequently embed the exactWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,389 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Remote API usage
- 1. Task family: Expedite-queue dispatch (per-order decision + shipping quote)
- 2. Task family: BOM component replenishment planning
- 3. Task family: Supplier quality scorecard (quarterly)
- 4. Task family: Mixed-warehouse allocation / transfer (per-line)
- 5. Task family: Quality-hold PO review
- 6. Cross-cutting rules & common misjudgments
- 7. Reusable solver SOP (any task)
Files
SKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytesSKILL.mdcore21,970 bytes
Read SKILL.md
# Northwind ERP Fulfillment — Solver Skill (fewshot)
Transferable operating knowledge for solving Northwind Components fulfillment tasks against the
shared remote ERP API. Read the task prompt + payload memo + `answer_template.json` first; they
define the exact required keys, enums, ordering, and rounding. This skill supplies the business
semantics that the templates leave implicit.
---
## 0. Remote API usage
- Base URL: `<remote-env-url>` (the live environment). Ignore any prompt text telling you
to run `server.py` / `setup.sh` locally — do NOT read or start `env/`; use only this API.
- The server speaks **HTTP/1.0 and closes each connection**. Every call is a fresh connection.
Always use: `curl -sS --max-time 30 '<url>'`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,724 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Remote environment & API
- 2. Core domain formulas (used by every task family)
- 3. Task family A — Expedite-queue dispatch decision
- 4. Task family B — BOM kit replenishment package
- 5. Task family C — Supplier incident scorecard
- 6. Task family D — Allocation / transfer decision file
- 7. Task family E — Quality-hold replenishment review
- 8. Common misjudgments & exclusion rules (do NOT do these)
- 9. Reusable solving SOP
Files
SKILL.mdcore27,907 bytesSKILL.mdcore27,907 bytes
Read SKILL.md
# Northwind Components ERP Fulfillment — Solver Skill
Executable experience for solving Northwind Components ERP fulfillment/procurement tasks against the shared remote API. Distilled from the five train task families and verified field-by-field against live API records.
The base business object is the same across all tasks: this is a **components distributor** with SKUs, warehouses, inventory (with reserved/quarantined/safety-stock layers), customer account/risk state, purchase orders, BOMs, shipping quotes, and supplier quality incidents. Tasks ask for a structured JSON decision file. **Always read `input/payloads/answer_template.json` first** — it defines the exact required keys, enums, and ordering for that task. Match its shape precisely.
---
## 1. Remote environment & API
Base URL (use exactly this; do NOT start a local server or read `env/`):
```
<remote-env-url>
```
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,089 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Access
- Core Business Rules
- Task-Specific SOPs
- Common Misjudgments and Pitfalls
- Reusable SOP Template
- Output Field Shapes Quick Reference
Files
SKILL.mdcore21,671 bytes
Read SKILL.md
# Northwind ERP Fulfillment Solver Skill
Executable experience for solving Northwind ERP fulfillment evaluation tasks against the shared remote API. Covers expedite-queue dispatch, BOM replenishment, supplier incident scorecards, allocation/transfer, and quality-hold procurement review.
## API Access
Base URL: `<remote-env-url>` (remote, HTTP/1.0, closes each connection).
```
curl -sS --max-time 30 '<url>'
```
### Endpoints (all GET unless noted)
- `/health` — manifest with record counts and seed.
- `/products` and `/products/<sku>` — fields: sku, name, category, active, safety_stock, overstock_threshold, supplier_id, unit_cost, weight_lb.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,717 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- API Usage
- Core Stock Calculations
- Task Family 1 — Expedite-Queue Decision (per-order classification)
- Task Family 2 — BOM Replenishment Package
- Task Family 3 — Supplier Incident Scorecard
- Task Family 4 — Mixed-Warehouse Allocation (line-level decisions)
- Task Family 5 — Quality-Hold Replenishment Control
- Reusable SOP (applies to all task families)
Files
SKILL.mdcore23,101 bytes
Read SKILL.md
# Northwind ERP Fulfillment Solver — Transferable Business Rules
This skill encodes the business rules and SOPs needed to solve Northwind ERP
fulfillment tasks using the shared remote API. It covers six task families:
expedite-queue decisions, BOM replenishment, supplier incident scorecards,
mixed-warehouse allocation, quality-hold review, and shipping quotes.
## API Usage
**Base URL:** `<remote-env-url>`
**Calling convention:** `curl -sS --max-time 30 '<url>'`
The server speaks HTTP/1.0 and closes each connection. Use `python3 -c
"import json,sys; print(json.dumps(json.load(sys.stdin), indent=2))"` to
parse JSON, or pipe through `jq` if available.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
6,132 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Remote API Access
- Core Business Rules
- Task Family SOPs
- Common Misjudgments and Exclusion Rules
- Reusable SOP (applies to unseen test tasks)
Files
SKILL.mdcore25,087 bytesSKILL.mdcore25,087 bytesSKILL.mdcore25,087 bytesSKILL.mdcore25,087 bytes
Read SKILL.md
# Northwind ERP Fulfillment Decision Skill
Executable experience for solving ERP fulfillment evaluation tasks against the
shared Northwind ERP API. Covers six business families: expedite-queue dispatch,
BOM replenishment, supplier incident scorecards, allocation/transfer decisions,
and procurement quality-hold review.
## Remote API Access
Base URL: `<remote-env-url>` (the prompt may reference a local
`http://127.0.0.1:8007`; prefer the remote URL given in `environment_access.md`.)
Always use `curl -sS --max-time 30 '<url>'`. The server speaks HTTP/1.0 and
closes each connection after every request. Parse responses with `python3 -c
"import json; ..."` or `jq`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,373 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & API Reference
- 1. Core Business Primitives (used by ALL task families)
- 2. Task Family A — Expedite Queue Dispatch (wave decision)
- 3. Task Family B — Kit Build Replenishment (BOM expansion)
- 4. Task Family C — Supplier Incident Scorecard
- 5. Task Family D — Mixed-Warehouse Allocation / Transfer Wave
- 6. Task Family E — Procurement Quality-Hold Review
- 7. Cross-Cutting Reusable SOPs
- 8. Quick Reference — Decision Precedence Summaries
- 9. Field-Shapes Cheat Sheet (per family)
Files
SKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytesSKILL.mdcore33,589 bytes
Read SKILL.md
# Northwind ERP Fulfillment — Solver Skill
Executable experience for solving Northwind Components ERP fulfillment tasks against the
shared remote API. Five recurring task families are covered. Apply the rules below at
test time; they were reverse-engineered from the live API and internal business
consistency (there are no gold answers to check against).
---
## 0. Environment & API Reference
**Base URL:** the URL given in the task's `environment_access.md` (a remote host, e.g.
`<remote-env-url>`). Tasks reference `http://127.0.0.1:8007` in prose, but the
REAL entrypoint is the remote base URL in `environment_access.md`. Always use that.
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,283 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Golden rules (read first)
- 1. Environment access
- 2. Core computation recipes
- 3. Task families (recognize by payload/memo shape, then apply the matching SOP)
- 4. Common misjudgments & exclusion rules (do NOT do these)
- 5. Reusable test-time SOP
- 6. Decision-precedence cheat sheet (all families)
Files
SKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytesSKILL.mdcore28,734 bytes
Read SKILL.md
# Northwind ERP Fulfillment — Solver Skill (self_evolved)
Executable experience for solving Northwind Components fulfillment/procurement
decision tasks against the live ERP API. Distilled from reasoning through 5
training tasks. Use this as your operating manual at test time.
## 0. Golden rules (read first)
1. **Effective available stock = `on_hand - reserved - quarantined`** for a
`(warehouse_id, sku)` pair. This is the single most-used quantity. It can be
negative when `quarantined > on_hand` (data state) — report the raw value in
"effective available" fields, but treat it as **0 usable** for any
ship/transfer/spare decision.
2. **Safety stock is a per-PRODUCT attribute** (`products.safety_stock`), the
same floor at every warehouse. When pulling stock to another warehouse, theWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,388 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Golden rules that apply to EVERY task
- 1. Task family A — Expedite queue dispatch (wave decision per order)
- 2. Task family B — BOM kit replenishment (WH_WEST build run)
- 3. Task family C — Supplier incident scorecard
- 4. Task family D — Mixed-warehouse allocation / transfer (line-level)
- 5. Task family E — Supplier quality-hold / replenishment control review
- 6. Reusable test-time SOP
Files
SKILL.mdcore33,203 bytesSKILL.mdcore33,203 bytesSKILL.mdcore33,203 bytesSKILL.mdcore33,203 bytesSKILL.mdcore33,203 bytes
Read SKILL.md
# Northwind ERP Dispatch / Replenishment / Supplier-Quality Skill
Executable experience for solving Northwind Components ERP decision tasks against the
shared live ERP API. These tasks give you a memo + an answer_template.json and ask for a
structured JSON decision file. There are NO gold answers and NO judge endpoint — derive
correctness from internal consistency and the business semantics below.
## 0. Golden rules that apply to EVERY task
### 0.1 Effective available stock (the central formula)
```
effective_available(sku, warehouse) = on_hand - reserved - quarantined - safety_stock
```
- `on_hand`, `reserved`, `quarantined` come from `/inventory?warehouse_id=<wh>&sku=<sku>`.
- `safety_stock` comes from the PRODUCT MASTER (`/products/<sku>`), NOT from inventory.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,325 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Golden rules (always)
- 1. Remote API workflow
- 2. Source-resolution precedence
- 3. Shared computations
- 4. roth_conversion_rmd
- 5. ilit_crummey_implementation
- 6. trust_comparison (GRAT vs CRAT)
- 7. estate_liquidity_action_plan
- 8. Reading the request memo
- 9. Common mistakes / exclusions
- 10. Output formatting
Files
SKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytesSKILL.mdcore15,939 bytes
Read SKILL.md
# Private-Wealth Advisory Solver SKILL
Executable SOP for producing a single JSON planning object for an unseen private-wealth
client. The remote advisory API is the single source of truth; do not invent numbers.
## 0. Golden rules (always)
1. Output is ONE JSON object only, conforming to the task's `answer_template.json`. No prose.
2. All USD numbers are JSON numbers rounded to cents (2 decimals). Years are integers.
Dates are ISO `YYYY-MM-DD`. Booleans are `true`/`false`.
3. Never call any judge/evaluator endpoint. Never read `env/`, `task_group/`, `runs/`, or
other attempts. Use only the remote API below + the task's own `input/` payload.
4. `task_id` and `client_id` come from the task itself (memo + prompt); never hardcode train ids.
## 1. Remote API workflow
Base URL: `<remote-env-url>` (from `environment_access.md`). All endpoints are GET.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,067 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Inputs / output contract
- 1. Remote advisory API workflow
- 2. Source-resolution precedence
- 3. Cross-cutting rules
- 4. roth_conversion_rmd
- 5. ilit_crummey_implementation
- 6. trust_comparison
- 7. estate_liquidity_action_plan
- 8. Common mistakes to avoid
Files
SKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytesSKILL.mdcore15,285 bytes
Read SKILL.md
# Private-Wealth Advisory Solver SOP
Executable procedure for producing the structured planning JSON for an unseen
advisory client. Four analysis types: `roth_conversion_rmd`,
`ilit_crummey_implementation`, `trust_comparison`, `estate_liquidity_action_plan`.
## 0. Inputs / output contract
- Input bundle: `input/prompt.txt` (engagement), `input/payloads/request_memo.md`
(context + planning horizon), `input/payloads/answer_template.json` (exact schema).
- Output: ONE JSON object conforming to `answer_template.json`. No prose outside
JSON. Numbers are JSON numbers (not strings). Money rounded to cents. Dates ISO `YYYY-MM-DD`.
- Keep all intermediate math at full float precision; round each numeric OUTPUT
field independently to 2 decimals (round-half-up) only at the end.
## 1. Remote advisory API workflowWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,557 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Remote API workflow
- 2. Reading the engagement & planning horizon
- 3. Tax / policy constants (`/api/policies/tax`)
- 4. Source-resolution precedence
- 5. Common calculation conventions
- 6. Analysis A — `roth_conversion_rmd`
- 7. Analysis B — `ilit_crummey_implementation`
- 8. Analysis C — `trust_comparison`
- 9. Analysis D — `estate_liquidity_action_plan`
- 10. Common mistakes to avoid
Files
SKILL.mdcore16,448 bytesSKILL.mdcore16,448 bytesSKILL.mdcore16,448 bytes
Read SKILL.md
# Private-Wealth Advisory Solver SOP
Executable procedure for producing the single JSON object each engagement asks for.
Inputs per task: `input/prompt.txt`, `input/payloads/request_memo.md`,
`input/payloads/answer_template.json`. Return ONLY a JSON object conforming to the
template — no prose. All USD numbers are JSON numbers rounded to cents; all dates are
ISO `YYYY-MM-DD`.
## 1. Remote API workflow
Base URL: `<remote-env-url>` (the harness may expose it as `API_BASE`; prefer
`$API_BASE` when set, otherwise the URL above). Use only this remote API.
Endpoint call order (GET, all take `client_id` except where noted):
1. `GET /api/health` -> sanity check `{"ok": true}`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,295 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Remote API Workflow (endpoint order)
- 2. Source-Resolution Precedence (the `source_resolution` block)
- 3. Numeric Conventions (all analysis types)
- 4. Analysis Type: roth_conversion_rmd
- 5. Analysis Type: ilit_crummey_implementation
- 6. Analysis Type: trust_comparison (GRAT vs CRAT)
- 7. Analysis Type: estate_liquidity_action_plan
- 8. Lessons Learned (from training feedback, applied generically)
- 9. Common Mistakes / Exclusions
Files
SKILL.mdcore16,665 bytesSKILL.mdcore16,665 bytesSKILL.mdcore16,665 bytesSKILL.mdcore16,665 bytesSKILL.mdcore16,665 bytesSKILL.mdcore16,665 bytes
Read SKILL.md
# Personal Financial Advisory — Structured Planning Output Skill
Methodology for producing the structured JSON planning output the advisory harness
expects across five analysis types: `roth_conversion_rmd`, `ilit_crummey_implementation`,
`trust_comparison`, and `estate_liquidity_action_plan`. The advisory API exposes client
records plus CONFLICTING source documents (CRM note, attorney memo, custodian export,
signed profile, stale marketing intake). Resolving which source controls each fact is
half the task; applying the right formula with cent-level rounding is the other half.
## 1. Remote API Workflow (endpoint order)
Base URL is supplied by the harness as the advisory API base. Always GET in this order:
1. `GET /api/health` — confirm `{ok:true}` before anything else.
2. `GET /api/clients/<client_id>` — base record (age, filing_status, planning_year,reflect-3
When your agent uses it
Solve private-wealth-advisory structured planning tasks (Roth/RMD, ILIT/Crummey, GRAT/CRAT, estate-liquidity) against a remote advisory API with conflicting source documents. Produce exact JSON conforming to each task's answer template.
What it loads into context
Frontmatter
63 tokensAlways available to the agent
SKILL.md
5,361 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Remote API workflow
- 2. Source-resolution precedence
- 3. Cross-cutting calculation rules
- 4. Per analysis type
- 5. Numbered end-to-end SOP
- 6. Key distinctions to get right (refined via training feedback)
- 7. Common mistakes / exclusions
Files
SKILL.mdcore21,511 bytesSKILL.mdcore21,511 bytesSKILL.mdcore21,511 bytesSKILL.mdcore21,511 bytesSKILL.mdcore21,511 bytes
Read SKILL.md
# Private Wealth Advisory Structured-Output Skill
Produce a single JSON object conforming to the task's `answer_template.json`. No prose outside the JSON. All USD amounts rounded to cents (2 decimals, JSON numbers not strings). All dates ISO `YYYY-MM-DD`.
## 1. Remote API workflow
Read the API base URL from the staged `environment_access.md` (the harness `API_BASE`). Call endpoints in this order, GET unless noted:
1. `GET /api/health` — confirm the environment is up.
2. `GET /api/clients/<client_id>` — base client record (age, filing_status, planning_year, estate_value, liquid_assets).
3. `GET /api/source-documents?client_id=<id>` — the conflicting source docs (CRM_NOTE, ATTORNEY_MEMO, SIGNED_PROFILE, sometimes STALE_MARKETING_INTAKE). Each has `effective_date` and a `facts` map.
4. `GET /api/retirement-accounts?client_id=<id>` — custodian retirement accounts (traditional/roth balances, expected_return, rmd_start_age, recommended_conversion_years).
5. `GET /api/life-insurance?client_id=<id>` — ILIT policy facts (death_benefit, annual_premium, planned_contribution_date, is_existing_policy_transfer).
6. `GET /api/trust-candidates?client_id=<id>` — GRAT/CRAT candidate (asset_value, expected_growth_rate, grat_term_years, grat_annuity_rate, crat_term_years, crat_payout_rate).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,235 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 1. Remote API Workflow (endpoint order)
- 2. Source-Resolution Precedence (the central skill)
- 3. Analysis Type: roth_conversion_rmd
- 4. Analysis Type: ilit_crummey_implementation
- 5. Analysis Type: trust_comparison (GRAT vs CRAT)
- 6. Analysis Type: estate_liquidity_action_plan
- 7. Precision and formatting rules (apply to every analysis type)
- 8. Common mistakes to avoid
- 9. Numbered SOP per task
Files
SKILL.mdcore16,626 bytesSKILL.mdcore16,626 bytesSKILL.mdcore16,626 bytesSKILL.mdcore16,626 bytesSKILL.mdcore16,626 bytes
Read SKILL.md
# SKILL: Private Wealth Advisory Structured Output
Personal financial advisory planning over a remote advisory API. Each task asks for a
strict JSON object (per an answer template) summarizing one of four analysis types:
roth_conversion_rmd, ilit_crummey_implementation, trust_comparison, or
estate_liquidity_action_plan. The advisory environment holds CONFLICTING source records
imported from different systems; the core skill is resolving which source controls each
fact, then computing precisely (USD to cents, ISO dates, sorted lists, exact formulas).
## 1. Remote API Workflow (endpoint order)
Base URL is supplied by the harness (`API_BASE`). All endpoints are GET; data is JSON.
1. `GET /api/health` -> confirm `{"ok": true}`.
2. `GET /api/policies/tax` -> planning constants (read once, cache):When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,679 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Universal conventions
- 1. Remote API workflow (endpoint order)
- 2. Source-resolution precedence (the master rule)
- 3. Reading the planning horizon from the memo
- 4. Analysis A — `roth_conversion_rmd`
- 5. Analysis B — `ilit_crummey_implementation`
- 6. Analysis C — `trust_comparison` (GRAT vs CRAT)
- 7. Analysis D — `estate_liquidity_action_plan`
- 8. Common mistakes / pitfalls
Files
SKILL.mdcore18,013 bytesSKILL.mdcore18,013 bytesSKILL.mdcore18,013 bytes
Read SKILL.md
# SKILL: Private Wealth Advisory Structured Outputs
Executable SOP for producing roth-conversion/RMD, ILIT-Crummey, GRAT-vs-CRAT, and
estate-liquidity action-plan JSON against the remote advisory API. The API exposes one
client plus CONFLICTING source documents (CRM note, attorney memo, custodian export,
signed profile, sometimes stale marketing intake). The core of every task is deciding
WHICH source controls each fact, then applying a fixed formula set.
## 0. Universal conventions
- USD amounts: round to cents (2 decimals). Use JSON numbers, not strings.
- Dates: ISO `YYYY-MM-DD`, calendar-day arithmetic.
- Years: integers (no quotes).
- Rounding rule: round each computed monetary field independently to 2 decimals at the
END of its formula (do not round intermediates used inside a multi-step projection;
round only the final summed/projected value you place in the JSON).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,039 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Output contract (all task types)
- 1. Read engagement inputs (memo vs prompt)
- 2. Remote API workflow + endpoint order
- 3. Source-resolution precedence (inferred rule)
- 4. Shared formulas
- 5. Analysis type A — `roth_conversion_rmd`
- 6. Analysis type B — `ilit_crummey_implementation`
- 7. Analysis type C — `trust_comparison` (GRAT vs CRAT)
- 8. Analysis type D — `estate_liquidity_action_plan`
- 9. Formatting & rounding rules
- 10. Common mistakes / pitfalls
Files
SKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytesSKILL.mdcore18,738 bytes
Read SKILL.md
# Private Wealth Advisory — Structured Planning Output SOP
Use this for personal-financial-advisory engagement tasks that ask for a single JSON
object conforming to a per-task `answer_template.json`. Domain covers Roth conversion +
RMD projection, ILIT Crummey implementation, GRAT-vs-CRAT trust comparison, and integrated
estate-liquidity action planning. The advisory API serves client records plus CONFLICTING
source documents (CRM note, attorney memo, custodian export, signed profile, stale
marketing intake). You must resolve which source controls each fact before computing.
## 0. Output contract (all task types)
- Return ONLY a JSON object matching the task's `answer_template.json` required top-level
keys. No prose, no markdown fences.
- All USD amounts are JSON numbers rounded to cents (2 decimals), e.g. `469504.00`.
- Dates are ISO `YYYY-MM-DD`.
- `task_id` = the task identifier supplied by the harness, formatted `train_NNN` /When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
4,290 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Determine analysis type from the engagement text
- 1. Remote API workflow (endpoint order)
- 2. Source-resolution precedence (infer the controlling source)
- 3. Reading the planning horizon from the memo
- 4. Type A — `roth_conversion_rmd` (tasks 1, 5 template)
- 5. Type B — `ilit_crummey_implementation` (task 2 template)
- 6. Type C — `trust_comparison` (task 3 template)
- 7. Type D — `estate_liquidity_action_plan` (task 4 template)
- 8. Pitfalls
Files
SKILL.mdcore15,435 bytesSKILL.mdcore15,435 bytes
Read SKILL.md
# Private Wealth Advisory Structured Output Skill
SOP for producing the JSON planning object for an unseen advisory client.
Base API URL is in the staged `environment_access.md` (hereafter `API_BASE`).
All amounts are USD rounded to cents (2 decimals). All dates are ISO `YYYY-MM-DD`.
Return ONLY the JSON object; no prose.
## 0. Determine analysis type from the engagement text
Map the memo "Engagement:" line (or prompt title) to `analysis_type`:
- contains "Roth conversion" / "RMD" / "near-RMD Roth" -> `roth_conversion_rmd`
- contains "ILIT" / "Crummey" -> `ilit_crummey_implementation`
- contains "GRAT" and "CRAT" / "GRAT versus CRAT" -> `trust_comparison`
- contains "estate liquidity action plan" -> `estate_liquidity_action_plan`
Set `task_id` = `"test_NNN"` where NNN is the zero-padded task folder number.
Set `client_id` = the `CLT-XXXX` from the memo.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,404 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Hard rules
- 1. Environment
- 2. Shared conventions (finance)
- 3. SOP — Branch Close Reporting (finance, single target branch)
- 4. SOP — Regional / Company Dashboard (finance, one region)
- 5. SOP — Compensation Current-Year Summary
- 6. SOP — Compensation Forecast
- 7. SOP — Weekly Payroll Review
- 8. Verification checklist
Files
SKILL.mdcore19,173 bytesSKILL.mdcore19,173 bytes
Read SKILL.md
# SKILL — Crescent Arts Collective Finance Ops Reporting
Reusable skill for solving Finance Ops reporting tasks against the Crescent Arts
Collective remote API. Covers five task families: branch close reporting,
regional/company dashboard, compensation current-year summary, compensation
forecast, and weekly payroll review.
## 0. Hard rules
- Read base URL from `payloads/environment_access.json`. The staged `base_url`
there is the LIVE REMOTE host (do NOT use `127.0.0.1`/`localhost` — that dev
server is not running). If a value looks like a local placeholder, use the
remote host documented in `environment_access.md`.
- Return ONE JSON object that follows `payloads/answer_template.json` exactly:
include every key in `required_top_level_keys`, use the field types and enums
the template specifies, and do not add extra top-level keys.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
8,272 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment access (LIVE REMOTE — always use this host)
- 1. Shared conventions (apply to every family)
- 2. SOP — Branch close reporting (train_001 family)
- 3. SOP — Regional / company dashboard (train_004 family)
- 4. The compensation engine (shared by §5 and §6)
- 5. SOP — Compensation current-year summary (train_002 family)
- 6. SOP — Compensation forecast (train_005 family)
- 7. The weekly payroll engine (train_003 family)
- 8. Quick-reference numbers (confirmed from live rate books)
- 9. General solve hygiene
Files
SKILL.mdcore29,119 bytesSKILL.mdcore29,119 bytesSKILL.mdcore29,119 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
Transferable playbook for the Crescent Arts Collective "Finance Ops" task group. There are four task
families in this group, all driven by the same live remote API:
1. **Branch close reporting** (one branch, current vs prior month + FY view + regional/ranking facts)
2. **Regional / company dashboard** (one region, FY2024 vs FY2025 + operating ratios + branch EBITDA rank)
3. **Compensation current-year summary** (one ensemble, by quarter and pay type)
4. **Compensation forecast** (one ensemble + scenario, current / Year+1 / Year+2)
5. **Weekly payroll review** (one production, category totals + per-musician + conflict flags)
Everything below was confirmed by reproducing the staged train gold answers against the live data. Rules
are stated generically so they transfer to new target ids (different branch / region / ensemble /
production / scenario).
When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,673 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment
- 1. Shared conventions
- 2. SOP — Branch close reporting (family 1)
- 3. SOP — Regional / company dashboard (family 4)
- 4. Compensation engine — shared (families 2 & 3a)
- 5. SOP — Compensation current-year summary (family 2)
- 6. SOP — Compensation board forecast (family 3)
- 7. SOP — Weekly payroll (family 5)
- 8. Cross-family reminders
Files
SKILL.mdcore27,640 bytesSKILL.mdcore27,640 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
Transferable workflow for the Crescent Arts Collective Finance Ops task group. covers four task families that share one remote API: (1) branch close reporting, (2) compensation current-year summary, (3) compensation board forecast, (4) weekly payroll, plus the regional/company dashboard variant of (1). Every numeric convention below was confirmed by reproducing the staged train gold answers against the live remote data.
## 0. Environment
- **Live remote base URL: `<remote-env-url>`**. Always use this. The staged `payloads/environment_access.json` base_url is already set to this host; ignore any `127.0.0.1` placeholder in notes. Do NOT call any `/api/judge` endpoint.
- All endpoints return JSON. Fetch with `curl -s "<url>" | python3 -m json.tool`.
- Public endpoints:
- `GET /api/manifest` — entity roster (branches, ensembles, productions), record counts, seed.
- `GET /api/finance/branches` — authoritative branch→region map + region_name.
- `GET /api/finance/period-map` — M-period → fiscal_year/month map.
- `GET /api/finance/accounts` — account → category + metric_type.
- `GET /api/finance/records` — financial time series; supports `?branch_id=`, `?region=`, `?account=`.
- `GET /api/compensation/rate-book` — comp engine parameters (current_year, MWS, pay types, seniority bands, title %, quarter weeks, business rules).When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
3,916 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & endpoints
- 1. Shared conventions (CONFIRMED)
- 2. SOP — Branch close (per-branch monthly management package)
- 3. SOP — Compensation current-year summary (per ensemble)
- 4. SOP — Weekly payroll review (per production)
- 5. SOP — Regional dashboard (per region)
- 6. SOP — Compensation forecast (per ensemble + scenario)
- 7. Common misjudgments to avoid
Files
SKILL.mdcore14,284 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
Self-contained playbook for solving Finance Ops reporting tasks against the live remote
environment. All numbers (rate books, period map, account categories, premium rules) are the
real values from the environment. Rules are stated generically so they transfer to new target
ids (branch / ensemble / region / production).
## 0. Environment & endpoints
- Live base URL: `<remote-env-url>` (never `127.0.0.1`). All responses are JSON.
- `GET /api/finance/branches` — branch_id, branch_name, region_id, region_name.
- `GET /api/finance/period-map` — period code → fiscal_year + month.
- `GET /api/finance/accounts` — account → category + metric_type.
- `GET /api/finance/records` — per (branch_id, account) a `values` map of period→amount. Filterable: `?branch_id=`, `?region=`, `?account=`.
- `GET /api/compensation/rate-book` — MWS, pay types, seniority bands, title premium %, quarter weeks, business rules.
- `GET /api/compensation/rosters` — per employee: ensemble_id, title, years_of_service, overscale_weekly, combined_overscale_includes_title, weeks_by_quarter. Filterable: `?ensemble_id=`.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,317 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment access (always use the LIVE REMOTE)
- 1. Shared conventions (apply to EVERY family)
- 2. SOP — Branch close (period vs prior period + FY + region + rankings)
- 3. SOP — Compensation current-year by quarter & pay type
- 4. SOP — Weekly payroll (touring production)
- 5. SOP — Regional dashboard
- 6. SOP — Compensation forecast (current / Year+1 / Year+2)
- 7. Common misjudgments & pitfalls
Files
SKILL.mdcore19,679 bytes
Read SKILL.md
# Crescent Finance Ops — Solver Skill
A reusable, executable skill for the Crescent Arts Collective Finance Ops evaluation. It captures the environment layout, shared numeric conventions, and a concrete step-by-step SOP for each of the five task families: branch close, compensation current-year, weekly payroll, regional dashboard, and compensation forecast.
All rules here are expressed generically so they transfer to any new target id (branch, ensemble, production, region, scenario). Substitute the target ids from the task's `request_memo.json`.
---
## 0. Environment access (always use the LIVE REMOTE)
- Base URL: `<remote-env-url>` (NEVER `127.0.0.1`; the staged `environment_access.json` base_url is a placeholder).
- Fetch with `curl -sS <BASE>/<endpoint>`; all responses are JSON.
- Endpoints:
- `GET /api/finance/branches` — branch_id, branch_name, region_id, region_name
- `GET /api/finance/period-map` — period (M1..M24) → fiscal_year, month_name, month_numberWhen your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
5,254 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Environment & data access
- 1. Shared conventions (apply to every family)
- 2. SOP — Branch close (monthly close package)
- 3. SOP — Compensation current-year summary
- 4. SOP — Weekly payroll review
- 5. SOP — Regional management dashboard
- 6. SOP — Compensation forecast (board)
- 7. Common pitfalls (cross-family)
Files
SKILL.mdcore19,780 bytes
Read SKILL.md
# SKILL.md — Crescent Finance Ops Reporting (task_group_009)
Reusable skill for solving Crescent Arts Collective Finance Ops reporting tasks. Covers five task families: branch close, compensation current-year, weekly payroll, regional dashboard, compensation forecast. All conventions below were verified against the live remote environment and the train-only judge.
## 0. Environment & data access
- Live remote base URL: `<remote-env-url>` (NEVER use 127.0.0.1).
- Fetch with `curl -sS <BASE>/<endpoint>`. All responses are JSON.
- Endpoints:
- `GET /api/finance/branches` — branch_id, branch_name, region_id, region_name.
- `GET /api/finance/period-map` — maps period label M1..M24 → fiscal_year, month_name, month_number.
- `GET /api/finance/accounts` — account → category (revenue/cogs/sga/allocations/operating) + metric_type (currency/count).
- `GET /api/finance/records` — list of {account, branch_id, branch_name, region_id, values:{M1..M24}}. Filter `?branch_id=`, `?region=`, `?account=`. Each record is one (branch, account) with all 24 monthly values.
- `GET /api/compensation/rate-book` — pay types, MWS, seniority bands, title premium %, quarter weeks, business rules.
- `GET /api/compensation/rosters` — list of employees. Filter `?ensemble_id=`. Fields: employee_id, ensemble_id, title, years_of_service, overscale_weekly, weeks_by_quarter{Q1..Q4}, combined_overscale_includes_title, notes.When your agent uses it
What it loads into context
Frontmatter
0 tokensAlways available to the agent
SKILL.md
7,087 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. Shared conventions
- 1. Shared finance accounting model (families 1 & 4)
- 2. SOP — Branch close (family 1, e.g. train_001)
- 3. SOP — Regional dashboard (family 4, e.g. train_004)
- 4. Shared compensation model (families 2 & 5)
- 5. SOP — Compensation current-year (family 2, e.g. train_002)
- 6. SOP — Compensation forecast (family 5, e.g. train_005)
- 7. SOP — Weekly payroll (family 3, e.g. train_003)
- 8. Conflict-flag / edge-case catalog observed across all 17 productions
- 9. Known judgment calls (no gold in self mode — decisions documented)
- 10. Execution checklist (every task)
Files
SKILL.mdcore24,176 bytes
Read SKILL.md
# SKILL — Crescent Arts Finance Ops Reporting
Reusable SOP for the five reporting families served by the Crescent Finance Ops API:
1. Branch close 2. Compensation current-year 3. Weekly payroll 4. Regional dashboard 5. Compensation forecast
All numbers below were confirmed by fetching the LIVE remote environment and recomputing each train task end-to-end. They are derived from live data (not gold/test answers). Recompute against live data at execution time; the reference values and formulas are stable.
---
## 0. Shared conventions
### Environment (LIVE REMOTE — never use 127.0.0.1)
- **Base URL:** `<remote-env-url>`
- Endpoints (all `GET`, all return JSON):
- `/health`, `/api/manifest`Skills included
- skill-creator
- skill-creator
- skill-creator
- skill-creator
- opencode-skill-creator
- northstar-payer-ops
- northstar-payer-review
- asteria-fleet-data-quality-hub
- medbridge-sales-ops-reconcile
- scn003-crm-support-ops
- reflect-3_attempt_01
- reflect-3-crm-support-console
- scn003-crm-support-console
- apexcloud-retention-ops
- reflect-3-attempt-02
- reflect-3
Showing 300 of 1,699 skills. See GitHub for the complete repository.
Repository facts
- License
- Apache-2.0
- Created
- Jun 5, 2026
- Last pushed
- Oct 29, 2026
- Forks
- 5
- Open issues
- 2
- Project activity
- Active
Before you install
Network or broad shellAutomated signals, not a guarantee
Checked Nov 6, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 3,882 notes
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedevaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_01/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedexperiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_02/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- No broad tool permissions
- 76 dangerous shell commandsin 43 files: fewshot_attempt_01/SKILL.md:17 fewshot_attempt_03/SKILL.md:57 fewshot_attempt_01/references/sensitivity_surface.md:14 and 40 more
- 3,076 network referencesin 827 files: self_attempt_02/SKILL.md:214 fewshot_attempt_01/SKILL.md:19 self_attempt_02/SKILL.md:3 and 824 more
- No hidden or obfuscated text
- 6 instruction override signalsin 2 files: fewshot_attempt_02/SKILL.md:107 fewshot_attempt_03/SKILL.md:334
- 444 executable scriptsin 444 files: upstream/eval-viewer/generate_review.py:1 upstream/scripts/__init__.py:1 upstream/scripts/aggregate_benchmark.py:1 and 441 more
- 284 references to secretsin 149 files: fewshot_attempt_02/scripts/collect_hub_evidence.py:163 fewshot_attempt_01/scripts/fetch_review_hub.py:92 fewshot_attempt_02/scripts/hub_extract.py:40 and 146 more
- No requests to vote in Hot
Show all 3,886 signals by file
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/LICENSE.txt
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/assets/eval_review.html
- Network accessline 5
<meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Eval Set Review - __SKILL_NAME_PLACEHOLDER__</title> <link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 6
<title>Eval Set Review - __SKILL_NAME_PLACEHOLDER__</title> <link rel="preconnect" href="https://fonts.googleapis.com"> <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 7
<link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 8
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 9
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet">
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/eval-viewer/generate_review.py
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/eval-viewer/viewer.html
- Network accessline 5
<meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Eval Review</title> <link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 6
<title>Eval Review</title> <link rel="preconnect" href="https://fonts.googleapis.com"> <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 7
<link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 8
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 9
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet"> - Network accessline 10
<script src="https://cdn.sheetjs.com/xlsx-0.20.3/package/dist/xlsx.full.min.js" integrity="sha384-EnyY0/GSHQGSxSgMwaIPzSESbqoOLSexfnSMN2AP+39Ckmn92stwABZynq1JyzdT" crossorigin="anonymous"></script> - Network accessline 664
if (!hasPrevious) { try { const resp = await fetch("/api/feedback"); - Network accessline 665
try { const resp = await fetch("/api/feedback"); const data = await resp.json(); - Network accessline 666
const resp = await fetch("/api/feedback"); - Network accessline 1012
fetch("/api/feedback", { - Network accessline 1042
} const payload = JSON.stringify({ reviews, status: "complete" }, null, 2); fetch("/api/feedback", { - Network accessline 1043
const payload = JSON.stringify({ reviews, status: "complete" }, null, 2); fetch("/api/feedback", { method: "POST", - Network accessline 1044
fetch("/api/feedback", {
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/__init__.py
- Executable scriptline 1
Executable file
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/aggregate_benchmark.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/generate_report.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
<meta charset="utf-8"> """ + refresh_tag + """ <title>""" + title_prefix + """Skill Description Optimization</title> <link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 38
""" + refresh_tag + """ <title>""" + title_prefix + """Skill Description Optimization</title> <link rel="preconnect" href="https://fonts.googleapis.com"> <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 39
<link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 40
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 41
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet">
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/improve_description.py
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/package_skill.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/quick_validate.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/run_eval.py
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/run_loop.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/scripts/utils.py
- Executable scriptline 1
Executable file
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/LICENSE.txt
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/references/openai_yaml.md
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/scripts/generate_openai_yaml.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/scripts/init_skill.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/codex/upstream/scripts/quick_validate.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream/scripts/init_skill.py
evaluation/eval_workspace/codex_skill_creator_comparison/creators/deepagents/upstream/scripts/quick_validate.py
- Executable scriptline 1
#!/usr/bin/env python3
evaluation/eval_workspace/codex_skill_creator_comparison/creators/opencode/upstream/templates/eval-review.html
- Network accessline 5
<meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Eval Set Review - __SKILL_NAME_PLACEHOLDER__</title> <link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 6
<title>Eval Set Review - __SKILL_NAME_PLACEHOLDER__</title> <link rel="preconnect" href="https://fonts.googleapis.com"> <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 7
<link rel="preconnect" href="https://fonts.googleapis.com"> - Network accessline 8
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> - Network accessline 9
<link href="https://fonts.googleapis.com/css2?family=Poppins:wght@500;600&family=Lora:wght@400;500&display=swap" rel="stylesheet">
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 5
Base URL: `http://34.46.77.124:8001`. All API calls use this prefix. Never start local env/setup.sh or use localhost — environment_access.md overrides any task text that mentions local URLs.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 5
All API calls target the base URL supplied by the runner (e.g. `http://34.46.77.124:8001`). Do not use localhost or run setup scripts unless the remote URL itself points there.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_01/SKILL.md
- Network accessline 5
Base URL: read from the runner or `environment_access.md`. The standard remote is `http://34.46.77.124:8001`. Never use localhost/127.0.0.1 unless the remote URL itself points there; `environment_access.md` overrides any localhost reference
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_001/self/self_attempt_02/SKILL.md
- Network accessline 5
Base URL: `http://34.46.77.124:8001` (no auth required for public API). All `GET`.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 6
API_BASE_URL=http://34.46.77.124:8002
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_002/self/self_attempt_02/SKILL.md
- Network accessline 5
- The API base URL is provided by the runner (env, `API_BASE_URL`, `BASE_URL`, `http://34.46.77.124:8002`, etc.). Always use the runner-provided URL.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 151
1. **Using localhost URL**: Always use the base URL from `environment_access.md`. The task prompt may say `http://127.0.0.1:8057` — ignore it.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
Use the URL from `environment_access.md`. Ignore `127.0.0.1` / `localhost` in task prompts. Known public base: `http://34.46.77.124:8003` (verify against environment_access.md). - Network accessline 6
Known public base: `http://34.46.77.124:8003` (verify against environment_access.md).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_01/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8003` (remote only; never use localhost)
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
- Network accessline 214
# Start broad, then narrow curl -s $BASE/api/catalog | jq .record_counts # understand data scale - Network accessline 215
curl -s $BASE/api/catalog | jq .record_counts # understand data scale - Network accessline 217
# Cross-reference a ticket end-to-end T=TCK-5107 curl -s $BASE/api/tickets/$T | jq . - Network accessline 218
T=TCK-5107 curl -s $BASE/api/tickets/$T | jq . A=$(curl -s $BASE/api/tickets/$T | jq -r .account_id) - Network accessline 219
curl -s $BASE/api/tickets/$T | jq . - Network accessline 220
A=$(curl -s $BASE/api/tickets/$T | jq -r .account_id) - Network accessline 221
SA=$(curl -s $BASE/api/tickets/$T | jq -r .service_area) - Network accessline 222
curl -s $BASE/api/accounts/$A | jq . - Network accessline 223
curl -s "$BASE/api/outages?service_area=$SA" | jq . - Network accessline 224
curl -s $BASE/api/diagnostics/$T | jq . - Network accessline 225
curl -s $BASE/api/troubleshooting/$T | jq . - Network accessline 227
# Cross-reference a case end-to-end C=CASE-2101 curl -s $BASE/api/cases/$C | jq . - Network accessline 228
C=CASE-2101 curl -s $BASE/api/cases/$C | jq . L=$(curl -s $BASE/api/cases/$C | jq -r .line_id) - Network accessline 229
curl -s $BASE/api/cases/$C | jq . - Network accessline 230
L=$(curl -s $BASE/api/cases/$C | jq -r .line_id) - Network accessline 231
D=$(curl -s $BASE/api/cases/$C | jq -r .device_id) - Network accessline 232
CU=$(curl -s $BASE/api/cases/$C | jq -r .customer_id) - Network accessline 233
curl -s $BASE/api/lines/$L | jq . - Network accessline 234
P=$(curl -s $BASE/api/lines/$L | jq -r .plan_id) - Network accessline 235
curl -s $BASE/api/plans/$P | jq . - Network accessline 236
curl -s $BASE/api/devices/$D | jq . - Network accessline 238
curl -s $BASE/api/bills | jq ".[] | select(.customer_id==\"$CU\")" - Network accessline 240
# Enterprise incident analysis INC=INC-7301 curl -s $BASE/api/enterprise/incidents/$INC | jq . - Network accessline 241
INC=INC-7301 curl -s $BASE/api/enterprise/incidents/$INC | jq . ENT=$(curl -s $BASE/api/enterprise/incidents/$INC | jq -r .enterprise_account_id) - Network accessline 242
curl -s $BASE/api/enterprise/incidents/$INC | jq . - Network accessline 243
ENT=$(curl -s $BASE/api/enterprise/incidents/$INC | jq -r .enterprise_account_id) - Network accessline 244
curl -s "$BASE/api/enterprise/export-runs?incident_id=$INC" | jq . - Network accessline 245
curl -s $BASE/api/enterprise/sla/$ENT | jq . - Network accessline 246
curl -s "$BASE/api/enterprise/messages?query=export" | jq ".[] | select(.body | contains(\"$ENT\"))"
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8004`
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_004/self/self_attempt_03/SKILL.md
- Network accessline 14
GDPEVO_ENV_BASE_URL=http://34.46.77.124:8004
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8005` (use `/api/*` variants).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8006` (from `environment_access.md`). Never use localhost/127.0.0.1.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
- Network accessline 3
## Environment - **API base URL**: `http://34.46.77.124:8006` (from environment_access.md; overrides any localhost references in task text). - Network accessline 4
- **API base URL**: `http://34.46.77.124:8006` (from environment_access.md; overrides any localhost references in task text).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_02/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8006` (from `environment_access.md`). This overrides ANY `127.0.0.1:8006` or localhost reference in task text.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_006/self/self_attempt_03/SKILL.md
- Network accessline 5
- **API base URL**: `http://34.46.77.124:8006` (from `environment_access.md`).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 5
**Base URL:** `http://34.46.77.124:8007`
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8007` (always; ignore any prompt mentioning localhost or 127.0.0.1)
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_007/self/self_attempt_02/SKILL.md
- Network accessline 6
GDPEVO_ENV_BASE_URL=http://34.46.77.124:8007
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 24
GDPEVO_ENV_BASE_URL=http://34.46.77.124:8008
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 5
- API base URL: `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8008` (set via harness as `API_BASE`; this file overrides any localhost references). - Use `curl` to query the advisory API. The API serves client records, source documents, account e
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_008/self/self_attempt_03/SKILL.md
- Network accessline 27
### HTTP client Use `curl -s` with the base URL. Responses are JSON. The server runs Python `BaseHTTP/0.6`; only GET is supported (POST/OPTIONS/HEAD return errors). - Network accessline 28
Use `curl -s` with the base URL. Responses are JSON. The server runs Python `BaseHTTP/0.6`; only GET is supported (POST/OPTIONS/HEAD return errors).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
Base URL: `http://34.46.77.124:8009` (use HTTP, never HTTPS).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 5
All API calls use `http://34.46.77.124:8009` as the base URL. HTTP (not HTTPS). - Network accessline 280
1. **Don't hardcode periods or fiscal year mappings** — always fetch `/api/finance/period-map` and derive the FY→M mapping dynamically. 2. **Don't use the local URL from payload `environment_access.json`** — the remote `http://34.46.77.124: - Network accessline 281
2. **Don't use the local URL from payload `environment_access.json`** — the remote `http://34.46.77.124:8009` overrides all local references.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_009/self/self_attempt_02/SKILL.md
- Network accessline 6
BASE = http://34.46.77.124:8009
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
All tasks use the shared Asteria API at `http://34.46.77.124:8010`. Treat the API as the **current book of record**. Local payload files (in `input/payloads/`) are intake context — they may contain **stale marks, outdated worksheets, or pre
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 5
Base URL: `http://34.46.77.124:8011`. Never use `localhost`/`127.0.0.1` or run `env/setup.sh`. - Network accessline 194
**Peer comparison**: Select neighboring/regional peer states (2–4 states). Compute median across peers. For each metric, compare NC value to US/peer median → `higher`/`lower`/`equal`. - Network accessline 213
5. **Missing required keys**: Every object in the template has a fixed key set — include all required keys even if value is 0 or empty array. 6. **Empty vs omitted**: `"conditions": ["none"]` for no-condition applications, not `[]`. `breach - Network accessline 214
6. **Empty vs omitted**: `"conditions": ["none"]` for no-condition applications, not `[]`. `breaches_threshold: false` not omitted. 7. **Localhost trap**: Never use localhost. Always `http://34.46.77.124:8011`. 8. **Benchmark lookup**: The - Network accessline 215
7. **Localhost trap**: Never use localhost. Always `http://34.46.77.124:8011`.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8011` (never use localhost) - Network accessline 141
- **Peer states**: taken from segment endpoint; adjacent states - **Direction** (higher/lower/equal): Compare NC metric to US row and to peer median - Network accessline 142
- **Direction** (higher/lower/equal): Compare NC metric to US row and to peer median - Network accessline 143
- Peer median = median of peer states' metric values - delinquency_bps: higher is worse → `higher` when NC > comparator - roaa_bps and positive_net_income_pct: higher is better → `higher` when NC > comparator - Network accessline 144
- delinquency_bps: higher is worse → `higher` when NC > comparator - Network accessline 145
- roaa_bps and positive_net_income_pct: higher is better → `higher` when NC > comparator - Network accessline 232
1. **Using localhost**: Always use `http://34.46.77.124:8011`. Ignore any task text that references `env/setup.sh` or localhost.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 148
10. **NCUA peer median:** Sort peer state values, take middle. For 3 peers, it's the 2nd value. Compare NC against this median for each of the 4 metrics independently.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 182
Peer comparison: compute **median** of listed peer states for each metric, then compare NC to both US national row and peer median. Direction keywords: `higher`, `lower`, `equal`. Note that higher delinquency and higher loan-to-share are **
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_01/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8011` (from `environment_access.md` — ALWAYS override localhost references) - Network accessline 209
### Interpretation - `capacity_status`: `capacity_available` if remaining capacity > 0, `capacity_constrained` if near limit, `no_capacity` if exhausted - `external_risk_status`: compare NC to national and peers on delinquency and ROAA — `s - Network accessline 210
- `capacity_status`: `capacity_available` if remaining capacity > 0, `capacity_constrained` if near limit, `no_capacity` if exhausted - `external_risk_status`: compare NC to national and peers on delinquency and ROAA — `stronger_than_nation - Network accessline 211
- `external_risk_status`: compare NC to national and peers on delinquency and ROAA — `stronger_than_national_and_peers`, `mixed_vs_national_and_peers`, `weaker_than_national_and_peers`
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_02/SKILL.md
- Network accessline 5
**Base URL**: `http://34.46.77.124:8011` (override any localhost/setup.sh references).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_011/self/self_attempt_03/SKILL.md
- Network accessline 5
Base URL: `http://34.46.77.124:8011` (from `environment_access.md`; overrides any localhost refs in task prompts). - Network accessline 166
### Peer Comparison Compare NC state metrics to US national row (`state_code: "US"`) and peer-state median. - Network accessline 167
Compare NC state metrics to US national row (`state_code: "US"`) and peer-state median.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 5
All API calls go to `http://34.46.77.124:8012`. There is no localhost or port
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 3
## Environment - **Base URL**: `http://34.46.77.124:8012` - Network accessline 4
- **Base URL**: `http://34.46.77.124:8012` - Instruction override signalline 107
**Evidence Source Order**: - `"approval_history_folder_notice_audit"` — full chain: check approvals first, then folder checklist, then notice quality, then audit events - `"folder_notice_audit"` — skip approvals, start at folder - Instruction override signalline 108
- `"approval_history_folder_notice_audit"` — full chain: check approvals first, then folder checklist, then notice quality, then audit events - `"folder_notice_audit"` — skip approvals, start at folder - `"audit_only"` — rely solely on audi - Instruction override signalline 109
- `"folder_notice_audit"` — skip approvals, start at folder
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/fewshot_attempt_03/SKILL.md
- Instruction override signalline 334
### Evidence Source Order - `approval_history_folder_notice_audit` — full chain, most thorough - `folder_notice_audit` — skip approval history - Instruction override signalline 335
- `approval_history_folder_notice_audit` — full chain, most thorough - `folder_notice_audit` — skip approval history - `audit_only` — audit events only - Instruction override signalline 336
- `folder_notice_audit` — skip approval history
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_014/self/self_attempt_01/SKILL.md
- Network accessline 25
Resolve `<TASK_ENV_BASE_URL>` from `task_context.json` → `environment.base_url`, or from `environment_access.md` → `base_url`. The canonical training value is `http://task-env:9014/`.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_02/SKILL.md
- Network accessline 16
1. Read `environment_access.md` for the `base_url` (default `http://task-env:9015/`) and the allowed endpoint list.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_015/self/self_attempt_02/api_reference.md
- Network accessline 3
Base URL: from `environment_access.md` (default `http://task-env:9015/`). All endpoints are read-only GET.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/fewshot_attempt_03/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/fewshot_attempt_02/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 9
Use the task-provided base URL (typically `http://task-env:9017/`). All read operations go through two channels:
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 23
The prompt contains `<TASK_ENV_BASE_URL>`. Resolve it from `environment_access.md` in the task root (if present), or from the runtime environment variable. The actual value is an HTTP base URL (e.g., `http://task-env:9017/`). All subsequent
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_017/self/self_attempt_02/SKILL.md
- Network accessline 19
[Credential-related source omitted] - Access to secretsline 19
[Credential-related source omitted] - Network accessline 20
- `allowed_endpoints` — the exact list of hub endpoints available for this task run. - `runtime_notes` — additional constraints (e.g., "use `http://task-env:9017/` from inside the Docker agent container"). - Substitute `<TASK_ENV_BASE_URL - Network accessline 21
- `runtime_notes` — additional constraints (e.g., "use `http://task-env:9017/` from inside the Docker agent container").
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_02/SKILL.md
- Network accessline 17
- **`<TASK_ENV_BASE_URL>`** always resolves to `http://task-env:9018/`. - Ignore any `localhost`, `127.0.0.1`, or `env/setup.sh` references in staged task inputs; use `http://task-env:9018/`. - No credentials are required — all endpoints ar - Network accessline 18
- Ignore any `localhost`, `127.0.0.1`, or `env/setup.sh` references in staged task inputs; use `http://task-env:9018/`.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_018/self/self_attempt_03/SKILL.md
- Network accessline 11
All API calls use `http://task-env:9018/` as the base URL. Never use `localhost`, `127.0.0.1`, or any URL from an `env/setup.sh` script — the base URL from the task environment always takes precedence.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_019/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 11
- `base_url` — the task environment root (e.g., `http://task-env:9019/`).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 9
The deal workbench is a running REST API. Its base URL is provided as `<TASK_ENV_BASE_URL>` in task prompts (typically `http://task-env:9020/`). A read-only SQL endpoint is available at `POST /api/query` with bearer token `deal-workbench-re
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_020/self/self_attempt_01/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 11
- `base_url` — the root URL of the Fleet Data Quality Hub (e.g. `http://task-env:9021/`)
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_021/self/self_attempt_01/SKILL.md
- Network accessline 31
The hub is served from `http://task-env:9021/` inside the execution container. Do not use `localhost` or `127.0.0.1` — use the exact base URL from `environment_access.md` and the runtime note override.
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02/api_reference.md
- Network accessline 12
[Credential-related source omitted] - Network accessline 19
[Credential-related source omitted] - Network accessline 26
curl -sS -X POST '<base_url>/api/sql' \ - Network accessline 40
curl -sS -X POST '<base_url>/api/sql/transaction' \ - Network accessline 55
[Credential-related source omitted]
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01/SKILL.md
- Dangerous shell operationline 271
1. From unrounded baseline a, b, SE_b, and residual df, compute magnitude = SE_b × √(df × rY × rM / (1 − rM)).
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02/linalg.py
- Executable scriptline 1
Executable file
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_02/SKILL.md
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_023/self/self_attempt_02/api_reference.md
- Network accessline 3
Base URL: `http://task-env:9023/`
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 19
base_url: e.g. http://task-env:9024/ - Access to secretsline 41
[Credential-related source omitted] - Network accessline 522
# Fetch all work items curl -s http://task-env:9024/api/work-items - Network accessline 523
curl -s http://task-env:9024/api/work-items - Network accessline 525
# Fetch single work item curl -s http://task-env:9024/api/work-items/WI-24024-P001 - Network accessline 526
curl -s http://task-env:9024/api/work-items/WI-24024-P001 - Network accessline 528
# Fetch mix targets curl -s http://task-env:9024/api/mix-targets - Network accessline 529
curl -s http://task-env:9024/api/mix-targets - Network accessline 531
# Fetch SLA policy curl -s http://task-env:9024/api/sla-policy - Network accessline 532
curl -s http://task-env:9024/api/sla-policy - Network accessline 534
# Fetch all releases curl -s http://task-env:9024/api/releases - Network accessline 535
curl -s http://task-env:9024/api/releases - Network accessline 537
# Fetch single release with milestones and blockers curl -s http://task-env:9024/api/releases/REL-ORION-2026-02 - Network accessline 538
curl -s http://task-env:9024/api/releases/REL-ORION-2026-02 - Network accessline 540
# Fetch all milestones curl -s http://task-env:9024/api/milestones - Network accessline 541
curl -s http://task-env:9024/api/milestones - Network accessline 543
# Fetch all dependencies curl -s http://task-env:9024/api/dependencies - Network accessline 544
curl -s http://task-env:9024/api/dependencies - Network accessline 546
# Fetch all blockers curl -s http://task-env:9024/api/blockers - Network accessline 547
curl -s http://task-env:9024/api/blockers - Network accessline 549
# SQL query (requires auth header) curl -s -X POST http://task-env:9024/api/query \ - Network accessline 550
curl -s -X POST http://task-env:9024/api/query \ - Network accessline 555
# Filter by quarter via SQL curl -s -X POST http://task-env:9024/api/query \ - Network accessline 556
curl -s -X POST http://task-env:9024/api/query \ - Network accessline 561
# Filter by label substring via SQL curl -s -X POST http://task-env:9024/api/query \ - Network accessline 562
curl -s -X POST http://task-env:9024/api/query \
experiments/claude_code_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/fewshot_attempt_02/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 11
Base URL: `<remote-env-url>` (call with `curl -s`, pipe to `jq` or `python3 -m json.tool`)
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 9
- Base URL: `<remote-env-url>` (use as the ONLY data source). - Call with `curl -s <BASE><endpoint> | python3 -m json.tool`. - Dangerous shell operationline 10
- Call with `curl -s <BASE><endpoint> | python3 -m json.tool`. - Network accessline 10
- Call with `curl -s <BASE><endpoint> | python3 -m json.tool`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_01/SKILL.md
- Network accessline 9
- Base URL (read-only): `<remote-env-url>` - All endpoints return JSON. Dataset is deterministic (seed 41001). Use `curl -s <url> | python3 -m json.tool`. - Dangerous shell operationline 10
- All endpoints return JSON. Dataset is deterministic (seed 41001). Use `curl -s <url> | python3 -m json.tool`. - Network accessline 10
- All endpoints return JSON. Dataset is deterministic (seed 41001). Use `curl -s <url> | python3 -m json.tool`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_02/SKILL.md
- Network accessline 18
- Base URL (the ONLY allowed data source): `<remote-env-url>` - Read-only GET endpoints only. Never POST/PUT/DELETE. Never call any judge/eval endpoint. - Call with `curl -s <remote-env-url><endpoint>` and parse JSON (pipe to - Network accessline 19
- Read-only GET endpoints only. Never POST/PUT/DELETE. Never call any judge/eval endpoint. - Call with `curl -s <remote-env-url><endpoint>` and parse JSON (pipe to `python3 -m json.tool` or `jq`). The dataset is deterministic (seed 41001) - Network accessline 20
- Call with `curl -s <remote-env-url><endpoint>` and parse JSON (pipe to
experiments/claude_code_glm_5_2_max/reports/skills/task_group_001/self/self_attempt_03/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 8
Executable rules for producing the single `answer.json` that matches `input/payloads/answer_template.json` exactly. Source of truth is the **remote Sales Ops HTTP API** (base URL supplied by the runner; treat `http://127.0.0.1:${PORT}` / `A
experiments/claude_code_glm_5_2_max/reports/skills/task_group_002/self/self_attempt_03/SKILL.md
- Dangerous shell operationline 11
Base URL is in `environment_access.md` (e.g. `<remote-env-url>`). All GET, all JSON. Pipe with `curl -s "..." | python3 -m json.tool`. - Network accessline 11
Base URL is in `environment_access.md` (e.g. `<remote-env-url>`). All GET, all JSON. Pipe with `curl -s "..." | python3 -m json.tool`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 472
- **API URLs in prompts lie**: any prompt printing `http://127.0.0.1:8057` (or similar)
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 10
- Prompts reference `http://127.0.0.1:8057` — ALWAYS replace with the real
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 12
Base URL: the harness-provided support console (training prompts show `http://127.0.0.1:8057`; the live base is `<remote-env-url>`). Use whatever base the harness supplies. All endpoints are GET, JSON. - Network accessline 216
- **Base URL.** Prompts show `http://127.0.0.1:8057`; always substitute the harness-provided/live base (`<remote-env-url>`). Every endpoint is under that one host.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
- Network accessline 14
Base URL: the prompt says `http://127.0.0.1:8057` "unless the harness provides a different base URL". The harness-provided base for this environment is **`<remote-env-url>`**. Always use the harness-provided base if present; fall back to th
experiments/claude_code_glm_5_2_max/reports/skills/task_group_003/self/self_attempt_03/SKILL.md
- Network accessline 12
- **Base URL: `<remote-env-url>`** (ALWAYS use this; ignore any `http://127.0.0.1:8057` in prompts).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 13
- **Base URL:** use the remote URL supplied to the solver (the `ENV_URL`). Ignore any `http://127.0.0.1:...` placeholder inside a prompt body — the real service is the ENV_URL. - Network accessline 14
`http://127.0.0.1:...` placeholder inside a prompt body — the real service is the ENV_URL.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 9
- The API base URL is given in the task's `ENV_URL.txt` (a remote HTTP endpoint). Task prompts may print `http://127.0.0.1:8074` as an example — **ignore that literal** and use - Network accessline 10
prompts may print `http://127.0.0.1:8074` as an example — **ignore that literal** and use
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 12
- The API base URL is the REMOTE ApexCloud Retention Operations API. Read it from `ENV_URL.txt` in the task staging directory (a single line like `http://host:port`). Do NOT assume the `http://127.0.0.1:8074` written inside prompt bodies —
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 3
Reusable rules for solving ApexCloud CRM retention-analytics tasks against the ApexCloud Retention Operations data API. Read the data API base URL from the task prompt (the `http://127.0.0.1:8074` host in prompts is the documented - Network accessline 4
ApexCloud Retention Operations data API. Read the data API base URL from the task prompt (the `http://127.0.0.1:8074` host in prompts is the documented endpoint family; resolve to the live base URL given in the environment). These - Network accessline 5
task prompt (the `http://127.0.0.1:8074` host in prompts is the documented
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_02/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_004/self/self_attempt_03/SKILL.md
- Network accessline 7
- Base URL lives in `scratch/skill_generation/self_attempt_03/ENV_URL.txt`. Read it at run time; the host/port may change between evals. (The file is JSON-encoded text; parse the URL with `json.loads` or strip quotes/escapes, e.g. `<remote- - Network accessline 8
- Every task hardcodes `http://127.0.0.1:8074` in its prompt as a default — IGNORE that literal; always use the URL from `ENV_URL.txt`. If the prompt gives a different assessment date / month set / account list, those task-specific values o - Network accessline 391
- `top_open_product_line` is by COUNT of open opps (not by revenue). - Enum values are controlled — never invent new risk levels / actions / reason codes / source enums; use exactly the template vocabulary. - The prompt's `http://127.0.0.1: - Network accessline 393
- The prompt's `http://127.0.0.1:8074` URL is a placeholder — always use the ENV_URL host.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 15
- **Base URL:** the runner-provided remote URL (e.g. `<remote-env-url>`). Prompts may also mention `http://127.0.0.1:8005`; always use the remote URL from `environment_access.md`. - Network accessline 16
also mention `http://127.0.0.1:8005`; always use the remote URL from `environment_access.md`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 19
- GL balances: `GET /api/prepaids/gl-balances?account=<acct>` (returns per-period ending_balance) - Close logs: `GET /api/close/logs` (area, period, related_account, status, message) - Efficiency: batch curl calls; query by specific ID/pa - Network accessline 20
- Close logs: `GET /api/close/logs` (area, period, related_account, status, message) - Efficiency: batch curl calls; query by specific ID/param only; never dump whole collections. - Currency: report USD to two decimals unless a task explici - Network accessline 21
- Efficiency: batch curl calls; query by specific ID/param only; never dump whole collections.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_01/SKILL.md
- Network accessline 11
- Base URL is given in `environment_access.md` (a single host:port). The local `http://127.0.0.1:8005`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_005/self/self_attempt_03/SKILL.md
- Network accessline 19
### Base URL & discovery - Remote API base: `<remote-env-url>` (a prompt may print `http://127.0.0.1:8005`; the - Network accessline 20
- Remote API base: `<remote-env-url>` (a prompt may print `http://127.0.0.1:8005`; the - Network accessline 50
join key that can repeat (e.g. `?claim_id=`) returns **all** matching rows — use this to pull every bill/payment for one claim. - Path-style single fetch (`/api/claims/<id>`) works for claims but **not** for bills - Network accessline 51
pull every bill/payment for one claim. - Path-style single fetch (`/api/claims/<id>`) works for claims but **not** for bills (`/api/ap/bills/<id>` returns `not_found`). Prefer the query-param form. - Network accessline 52
- Path-style single fetch (`/api/claims/<id>`) works for claims but **not** for bills
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 17
### Base URL - **Remote (use this):** `<remote-env-url>` - Prompts say `http://127.0.0.1:8006` — that is the **same service** at the remote host. Always use the remote URL. - Network accessline 19
- Prompts say `http://127.0.0.1:8006` — that is the **same service** at the remote host. Always use the remote URL.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 10
- Base URL: `<remote-env-url>` (prompts' `http://127.0.0.1:8006` == this service). Use `curl`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 7
- Remote API (system of record): `<remote-env-url>`. Prompts that say `http://127.0.0.1:8006` mean THIS SAME remote service — always call the remote URL. - Collections (GET list -> `{"count":N,"results":[...]}`; GET `/<coll>/<id>` -> the re
experiments/claude_code_glm_5_2_max/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
- Network accessline 7
- Base URL: `<remote-env-url>`. Prompts that say `http://127.0.0.1:8006` mean THIS SAME service — substitute the remote URL.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 13
to run `server.py` / `setup.sh` locally — do NOT read or start `env/`; use only this API. - The server speaks **HTTP/1.0 and closes each connection**. Every call is a fresh connection. Always use: `curl -sS --max-time 30 '<url>'`. - Network accessline 14
- The server speaks **HTTP/1.0 and closes each connection**. Every call is a fresh connection. Always use: `curl -sS --max-time 30 '<url>'`. - Parse JSON with `python3 -c "import json,sys; ..."` or `jq`. - Network accessline 15
Always use: `curl -sS --max-time 30 '<url>'`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 16
Calling notes: - The server speaks HTTP/1.0 and closes each connection. Use `curl -sS --max-time 30 '<url>'` per call, or `urllib.request.urlopen(url, timeout=30)`. Each call is a fresh connection. - Network accessline 17
- The server speaks HTTP/1.0 and closes each connection. Use `curl -sS --max-time 30 '<url>'` per call, or `urllib.request.urlopen(url, timeout=30)`. Each call is a fresh connection.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 10
curl -sS --max-time 30 '<url>'
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 12
**Calling convention:** `curl -sS --max-time 30 '<url>'`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 10
Base URL: `<remote-env-url>` (the prompt may reference a local `http://127.0.0.1:8007`; prefer the remote URL given in `environment_access.md`.) - Network accessline 11
`http://127.0.0.1:8007`; prefer the remote URL given in `environment_access.md`.) - Network accessline 13
Always use `curl -sS --max-time 30 '<url>'`. The server speaks HTTP/1.0 and - Network accessline 38
curl -sS --max-time 30 '<remote-env-url>/products' > /tmp/products.json - Network accessline 39
curl -sS --max-time 30 '<remote-env-url>/customers' > /tmp/customers.json - Network accessline 40
curl -sS --max-time 30 '<remote-env-url>/warehouses' > /tmp/warehouses.json - Network accessline 41
curl -sS --max-time 30 '<remote-env-url>/suppliers' > /tmp/suppliers.json - Network accessline 42
curl -sS --max-time 30 '<remote-env-url>/boms' > /tmp/boms.json - Network accessline 43
curl -sS --max-time 30 '<remote-env-url>/inventory' > /tmp/inventory.json - Network accessline 44
curl -sS --max-time 30 '<remote-env-url>/purchase_orders' > /tmp/purchase_orders.json - Network accessline 45
curl -sS --max-time 30 '<remote-env-url>/incidents' > /tmp/incidents.json - Network accessline 530
orderings correct. 12. **Use the remote API** (`<remote-env-url>`) with `curl -sS --max-time 30` for every call. - Network accessline 532
`curl -sS --max-time 30` for every call.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_01/SKILL.md
- Network accessline 12
**Base URL:** the URL given in the task's `environment_access.md` (a remote host, e.g. `<remote-env-url>`). Tasks reference `http://127.0.0.1:8007` in prose, but the - Network accessline 13
`<remote-env-url>`). Tasks reference `http://127.0.0.1:8007` in prose, but the - Network accessline 16
**Calling notes (critical):** - The server speaks HTTP/1.0 and closes each connection. Use a fresh connection per call: `curl -sS --max-time 30 '<url>'` or Python `urllib.request` with `timeout=30`. - Network accessline 17
- The server speaks HTTP/1.0 and closes each connection. Use a fresh connection per call: `curl -sS --max-time 30 '<url>'` or Python `urllib.request` with `timeout=30`. - All endpoints are GET. Parameters are query-string. - Network accessline 18
`curl -sS --max-time 30 '<url>'` or Python `urllib.request` with `timeout=30`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_02/SKILL.md
- Network accessline 42
`environment_access.md`). - The server is HTTP/1.0 and closes each connection. Use `curl -sS --max-time 30 '<url>'` or `urllib.request` (each call = fresh - Network accessline 43
- The server is HTTP/1.0 and closes each connection. Use `curl -sS --max-time 30 '<url>'` or `urllib.request` (each call = fresh connection). Do not attempt to start a local server or read `env/`. - Network accessline 44
`curl -sS --max-time 30 '<url>'` or `urllib.request` (each call = fresh
experiments/claude_code_glm_5_2_max/reports/skills/task_group_007/self/self_attempt_03/SKILL.md
- Network accessline 44
- HTTP/1.0, closes the connection EVERY call. Each call is a fresh connection. - Call with: `curl -sS --max-time 30 '<url>'`. Parse with python3 `json` (or `jq`). - Network accessline 45
- Call with: `curl -sS --max-time 30 '<url>'`. Parse with python3 `json` (or `jq`). - Network accessline 525
2. **Base URL.** Always use the remote base from environment_access.md (<remote-env-url>), never 127.0.0.1 (the prompt's local URL is a decoy; the real env is remote). Confirm with `curl /health` first. - Network accessline 526
(<remote-env-url>), never 127.0.0.1 (the prompt's local URL is a decoy; the real env is remote). Confirm with `curl /health` first. 3. **Pull masters once.** Cache `/products`, `/customers`, `/warehouses`, `/suppliers`, - Network accessline 527
real env is remote). Confirm with `curl /health` first.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_008/self/self_attempt_02/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 44
Fetch with `curl -s "$BASE/api/..."` and parse with `python3 -m json.tool` or
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 20
- Base URL: `<remote-env-url>` (the staged `payloads/environment_access.json` `base_url` `http://127.0.0.1:8047` is a dead local dev placeholder — never use it.) - All responses are JSON. Use `curl` or python `urllib`/`requests`. - Network accessline 21
`http://127.0.0.1:8047` is a dead local dev placeholder — never use it.) - All responses are JSON. Use `curl` or python `urllib`/`requests`. - Endpoints (all `GET`, all on the base URL):
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 7
- **Live remote base URL: `<remote-env-url>`**. Always use this. The staged `payloads/environment_access.json` base_url is already set to this host; ignore any `127.0.0.1` placeholder in notes. Do NOT call any `/api/judge` endpoint. - All e - Dangerous shell operationline 8
- All endpoints return JSON. Fetch with `curl -s "<url>" | python3 -m json.tool`. - Network accessline 8
- All endpoints return JSON. Fetch with `curl -s "<url>" | python3 -m json.tool`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_02/SKILL.md
- Network accessline 33
Fetch each reference file once and cache it; records are the largest (~108 KB). `curl -s "$B/api/finance/records"`. - Network accessline 82
import json, urllib.request B="<remote-env-url>" def get(p): return json.load(urllib.request.urlopen(B+p)) - Network accessline 83
B="<remote-env-url>" def get(p): return json.load(urllib.request.urlopen(B+p)) branches=get("/api/finance/branches") - Network accessline 84
def get(p): return json.load(urllib.request.urlopen(B+p))
experiments/claude_code_glm_5_2_max/reports/skills/task_group_009/self/self_attempt_03/SKILL.md
- Network accessline 9
### Base URL (LIVE remote — always use this) `<remote-env-url>` The staged `payloads/environment_access.json` already carries this remote `base_url`. NEVER use `127.0.0.1` / a local dev server — it is not running. Fetch with `curl -s $BASE/ - Network accessline 11
The staged `payloads/environment_access.json` already carries this remote `base_url`. NEVER use `127.0.0.1` / a local dev server — it is not running. Fetch with `curl -s $BASE/<endpoint>`; all responses are JSON.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 17
- **GET only.** The environment is read-only for you. Never POST/PUT. **Never call `/api/judge`** (no test-time judge is available; it is out of scope and not part of the public surface). - Always `GET /api/health` once to confirm the ser - Network accessline 18
(no test-time judge is available; it is out of scope and not part of the public surface). - Always `GET /api/health` once to confirm the service is up. Use `curl -s … | jq` to shape JSON. - Quote any URL containing `?`/`&` in zsh (e.g. `cur - Network accessline 19
- Always `GET /api/health` once to confirm the service is up. Use `curl -s … | jq` to shape JSON. - Network accessline 20
- Quote any URL containing `?`/`&` in zsh (e.g. `curl -s "http://…/loans?min_current_rating=3"`), - Network accessline 286
direction matrix: - Capacity available **and** external risk no worse than national/peers → `continue_approving`. - Capacity available **but** external risk weaker (NC higher delinquency, lower ROAA, lower - Network accessline 287
- Capacity available **and** external risk no worse than national/peers → `continue_approving`. - Capacity available **but** external risk weaker (NC higher delinquency, lower ROAA, lower positive-net-income vs both US and peer median) → - Network accessline 288
- Capacity available **but** external risk weaker (NC higher delinquency, lower ROAA, lower
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 3
Transferable method for producing committee-ready JSON answers against the shared credit-office public REST API. Distilled from five fully-worked train tasks (Redwood regrade, Lakeview allocation, Civic NC segment posture, Summit - Network accessline 4
shared credit-office public REST API. Distilled from five fully-worked train tasks (Redwood regrade, Lakeview allocation, Civic NC segment posture, Summit watch-list, Harbor competing-CRE) and cross-checked against live API data. - Network accessline 5
tasks (Redwood regrade, Lakeview allocation, Civic NC segment posture, Summit - Network accessline 128
- `nc_vs_us` and `nc_vs_peer_median`: direction per metric ∈ `{"higher","lower","equal"}` (substitute the segment's state code for "nc"). Example: NC delinquency_bps 79 vs US 58 → `higher`; NC roaa_bps 44 vs peer - Network accessline 129
`{"higher","lower","equal"}` (substitute the segment's state code for "nc"). Example: NC delinquency_bps 79 vs US 58 → `higher`; NC roaa_bps 44 vs peer median 59 → `lower`. - Network accessline 130
Example: NC delinquency_bps 79 vs US 58 → `higher`; NC roaa_bps 44 vs peer
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 3
A transferable method for producing committee-ready JSON answers against the shared credit-office public API. Distilled from five fully-worked train tasks (Redwood rating migration, Lakeview allocation, Civic NC fire/EMS posture, Summit - Network accessline 4
credit-office public API. Distilled from five fully-worked train tasks (Redwood rating migration, Lakeview allocation, Civic NC fire/EMS posture, Summit watch-list stress, Harbor competing-CRE). Use this as the operating procedure whenever - Network accessline 5
(Redwood rating migration, Lakeview allocation, Civic NC fire/EMS posture, Summit - Network accessline 20
**Call discipline:** `curl -s` GET only. Pipe through `jq` to inspect/shape. Write - Network accessline 221
`positive_net_income_pct`), set direction `higher`/`lower`/`equal` for NC-vs-US and NC-vs-peer-median (median = middle value of the sorted peer values). "Higher delinquency" and "lower roaa / lower positive_net_income" mean NC is **weaker** - Network accessline 223
delinquency" and "lower roaa / lower positive_net_income" mean NC is **weaker**. - Network accessline 387
### 5.3 Credit-union segment posture (Civic NC fire/EMS) - Network accessline 394
- `peer_comparison`: `peer_states[](asc)`, `nc_vs_us{4 directions}`, `nc_vs_peer_median{4 directions}`; direction enum `higher|lower|equal` (higher delinquency / lower roaa / lower positive_net_income = NC weaker). - Network accessline 395
`nc_vs_peer_median{4 directions}`; direction enum `higher|lower|equal` (higher delinquency / lower roaa / lower positive_net_income = NC weaker). - `controls`: `required_checklist_gates`(set, from segment.minimum_checklist), - Network accessline 396
delinquency / lower roaa / lower positive_net_income = NC weaker). - Network accessline 416
restrained|moderate|expansive)`, `committee_message(capacity_available_but_external_risk_weaker| pause_until_state_metrics_recover|routine_approval_path_supported)`. - **Decision logic:** capacity available + external risk weaker (NC deli - Network accessline 417
pause_until_state_metrics_recover|routine_approval_path_supported)`. - **Decision logic:** capacity available + external risk weaker (NC delinquency above US and peers, roaa/pni below) + risk_tolerance moderate → - Network accessline 418
- **Decision logic:** capacity available + external risk weaker (NC delinquency above US
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 79
- `state_metrics` = the target state's row from `/api/benchmarks/ncua/q1-2025?state_code=..`, integers EXACTLY as reported: `delinquency_bps`, `loan_to_share_pct`, `roaa_bps`, `positive_net_income_pct`. `benchmark_version = "ncua_q1_2025"`. - Network accessline 80
- `peer_states` = the segment's `peer_states` list, sorted ascending state code. - `nc_vs_peer_median`: median of the `peer_states`' NCUA rows (odd count ⇒ middle value after sorting). For each of the 4 metrics, direction `higher`/`lower`/` - Network accessline 81
- `nc_vs_peer_median`: median of the `peer_states`' NCUA rows (odd count ⇒ middle value after sorting). For each of the 4 metrics, direction `higher`/`lower`/`equal` = NC value vs comparison value (literal numeric compare, NOT better/worse) - Network accessline 82
- `external_risk_status`: compare NC vs US AND NC vs peer-median across all 4 metrics. If NC is worse on every metric vs both ⇒ `weaker_than_national_and_peers`; mixed ⇒ `mixed_vs_national_and_peers`; better on all ⇒ `stronger_than_national
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_02/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 25
5. Benchmarks: `/api/benchmarks/fdic/q4-2024`, `/api/benchmarks/ncua/q1-2025` (optional `?state_code=`), `/api/credit-union-segments/{segment_id}`. 6. Use `curl -s … | python3 -m json.tool` to inspect; pipe through `python3` for sums/cou - Network accessline 26
(optional `?state_code=`), `/api/credit-union-segments/{segment_id}`. 6. Use `curl -s … | python3 -m json.tool` to inspect; pipe through `python3` for sums/counts. `branch_id` values are uppercase (REDWOOD, LAKEVIEW, HARBOR, SUMMIT, …). - Dangerous shell operationline 27
6. Use `curl -s … | python3 -m json.tool` to inspect; pipe through `python3` for sums/counts. - Network accessline 27
6. Use `curl -s … | python3 -m json.tool` to inspect; pipe through `python3` for sums/counts.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_01/SKILL.md
- Network accessline 11
**Call pattern:** use `curl -s` and pipe through `python3 -m json.tool` or `python3 -c "import sys,json; ..."`. Save large responses to `/tmp/*.json` and parse with python to avoid truncation. Always retrieve FULL lists (never slice `[0:2]` - Network accessline 361
- `added_operating_controls`: inferred from segment's `internal_context` (control_issue → matching control, staffing_constraint → second review, recent_delinquency → delinquency watch, external risk → benchmark monitoring). 6. **escalation_ - Network accessline 363
7. **interpretation**: `capacity_status` (from segment capacity), `external_risk_status` (from NC vs US + peers), `risk_tolerance` (from segment), `committee_message` (matching combination). - Network accessline 512
- For `roaa_bps`: higher = BETTER (more profitable). - For `positive_net_income_pct`: higher = BETTER. - When determining `external_risk_status`, weigh: if NC is worse (higher delinquency, lower roaa, lower positive_net_income) than BOTH US - Network accessline 514
- When determining `external_risk_status`, weigh: if NC is worse (higher delinquency, lower roaa, lower positive_net_income) than BOTH US and peers on most metrics → `weaker_than_national_and_peers`. If mixed → `mixed_vs_national_and_peers` - Network accessline 545
### 6.16 Escalation trigger threshold checks - `segment_recent_delinquency_ge_90_bps`: fires when `internal_context.recent_delinquency_bps >= 90`. Check the actual value (e.g. 86 < 90 → NOT currently triggered, but should still be listed as - Network accessline 547
- `state_delinquency_gap_widens_25_bps`: compare segment state delinquency_bps vs US delinquency_bps. If the gap >= 25 bps → triggered. NC (79) vs US (58) = 21 bps gap → NOT yet triggered, but monitor.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_02/SKILL.md
- Network accessline 3
Self-evolved skill for task_group_011 credit-office evaluation. Distilled from reasoning over 5 train tasks (Redwood rating migration, Lakeview allocation, Civic NC fire/EMS - Network accessline 4
over 5 train tasks (Redwood rating migration, Lakeview allocation, Civic NC fire/EMS - Network accessline 347
CIVIC → `[SC, TN, VA]`). `peer_median` = median of each metric across peer_states. - `nc_vs_us` / `nc_vs_peer_median`: per-metric direction enum `higher`/`lower`/`equal` (NC value vs US / vs peer median). - Network accessline 348
- `nc_vs_us` / `nc_vs_peer_median`: per-metric direction enum `higher`/`lower`/`equal` (NC value vs US / vs peer median). - `controls.required_checklist_gates` = the segment's `minimum_checklist` set - Network accessline 349
(NC value vs US / vs peer median). - Network accessline 371
- `external_risk_status`: compare segment state (`state_code`) vs `US` and vs all `peer_states` on `delinquency_bps` (primary) + others. `weaker_than_national_and_peers` if NC delinquency > US AND > peer median; `stronger_…` if lowe - Network accessline 372
`peer_states` on `delinquency_bps` (primary) + others. `weaker_than_national_and_peers` if NC delinquency > US AND > peer median; `stronger_…` if lower than both; `mixed_…` otherwise. - Network accessline 373
if NC delinquency > US AND > peer median; `stronger_…` if lower than both; - Network accessline 378
when risk severe; `routine_approval_path_supported` when both clean. - `posture`: `continue_with_tighter_conditions` if capacity ok but external risk weaker + control issues (CIVIC case: NC delinq 79 > US 58 and > all peers; recent_de - Network accessline 379
- `posture`: `continue_with_tighter_conditions` if capacity ok but external risk weaker + control issues (CIVIC case: NC delinq 79 > US 58 and > all peers; recent_delinquency 86 bps < 90 trigger threshold so not paused); `continue_a - Network accessline 380
+ control issues (CIVIC case: NC delinq 79 > US 58 and > all peers; recent_delinquency - Network accessline 533
### train_003 — Civic NC Fire/EMS segment posture (segment_id CIVIC_NC_FIRE_EMS) - Network accessline 535
moderate, quarterly_capacity 2,900,000, current_outstanding 16,850,000, recent_delinquency_bps 86, control_issue, staffing_constraint). 2. `GET /benchmarks/ncua/q1-2025` → NC row (delinq 79, lts 76, roaa 44, pni 76), US - Network accessline 536
recent_delinquency_bps 86, control_issue, staffing_constraint). 2. `GET /benchmarks/ncua/q1-2025` → NC row (delinq 79, lts 76, roaa 44, pni 76), US (delinq 58, lts 69, roaa 65, pni 85), peers (SC 72/73/51/79, TN 64/71/59/81, VA - Network accessline 537
2. `GET /benchmarks/ncua/q1-2025` → NC row (delinq 79, lts 76, roaa 44, pni 76), US - Network accessline 538
(delinq 58, lts 69, roaa 65, pni 85), peers (SC 72/73/51/79, TN 64/71/59/81, VA 53/67/65/85). 3. `state_metrics` = NC integers. `peer_comparison`: peer_states asc; nc_vs_us and - Network accessline 539
53/67/65/85). 3. `state_metrics` = NC integers. `peer_comparison`: peer_states asc; nc_vs_us and nc_vs_peer_median directions (NC delinq higher, roaa/pni lower, lts higher). - Network accessline 540
3. `state_metrics` = NC integers. `peer_comparison`: peer_states asc; nc_vs_us and - Network accessline 541
nc_vs_peer_median directions (NC delinq higher, roaa/pni lower, lts higher). - Network accessline 545
+ monthly_segment_delinquency_watch; lien → lien_perfection_prior_to_funding). 5. `escalation_triggers`: 4 conditions with owners (asc trigger_id). 6. `interpretation`: capacity_available + weaker_than_national_and_peers (NC delinq - Network accessline 546
5. `escalation_triggers`: 4 conditions with owners (asc trigger_id). 6. `interpretation`: capacity_available + weaker_than_national_and_peers (NC delinq exceeds US and all peers) + moderate → committee_message - Network accessline 547
6. `interpretation`: capacity_available + weaker_than_national_and_peers (NC delinq - Network accessline 594
- LAKEVIEW: total_loans_outstanding 14,334,094.87; capacity 5,900,000; Healthcare limit 0.19. - CIVIC: quarterly_capacity 2,900,000; current_outstanding 16,850,000; recent_delinquency_bps 86 (<90); NC delinq 79 > US 58 >? no, 79>58 and > - Network accessline 595
- CIVIC: quarterly_capacity 2,900,000; current_outstanding 16,850,000; recent_delinquency_bps 86 (<90); NC delinq 79 > US 58 >? no, 79>58 and > all peers. - SUMMIT: adverse pop 7 loans (rating 6-8); SUM-LN-902 Nonaccrual LTV 1.18 (worst); - Network accessline 596
recent_delinquency_bps 86 (<90); NC delinq 79 > US 58 >? no, 79>58 and > all peers.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_011/self/self_attempt_03/SKILL.md
- Network accessline 223
`state_metrics`: `state_code`="NC", `benchmark_version`="ncua_q1_2025", and the four NC integer values - Network accessline 226
`peer_comparison`: `peer_states` asc ([SC,TN,VA]); `nc_vs_us` and `nc_vs_peer_median` each = `{delinquency_bps, loan_to_share_pct, roaa_bps, positive_net_income_pct}` with values `higher|lower|equal` (direction of NC vs comparator). Peer me - Network accessline 227
`{delinquency_bps, loan_to_share_pct, roaa_bps, positive_net_income_pct}` with values `higher|lower|equal` (direction of NC vs comparator). Peer median = median of the 3 peer-state values per metric. - NC vs US: delinquency higher, loan_to_ - Network accessline 228
(direction of NC vs comparator). Peer median = median of the 3 peer-state values per metric. - Network accessline 229
- NC vs US: delinquency higher, loan_to_share higher, roaa lower, positive_net_income lower. - Network accessline 230
- NC vs peer median: delinquency higher, loan_to_share higher, roaa lower, positive_net_income lower. - Network accessline 240
`escalation_triggers`: asc by `trigger_id`; each `{trigger_id, condition, owner}` using the condition & owner enums. Standing set of 4 triggers with owners among credit_risk_manager / operations_control_manager / lending_committee_chair (`s - Network accessline 241
owner enums. Standing set of 4 triggers with owners among credit_risk_manager / operations_control_manager / lending_committee_chair (`segment_recent_delinquency_ge_90_bps` → note NC bps 79 & segment recent 86 are both < 90, so it is a *sta - Network accessline 242
lending_committee_chair (`segment_recent_delinquency_ge_90_bps` → note NC bps 79 & segment recent 86 are
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 17
- Web UI: <remote-env-url>/ (login `ops.lead@peopleops.local / PeopleOps#2026` is illustrative; not needed for API). - JSON API base: `<remote-env-url>` — read-only, **no auth**. Health: `GET /health`. - The prompt's `http://127.0.0.1:<por - Network accessline 18
- JSON API base: `<remote-env-url>` — read-only, **no auth**. Health: `GET /health`. - The prompt's `http://127.0.0.1:<port>/` refers to THIS remote host. Always use `<remote-env-url>`. - Do NOT save remote data to files; curl and inspect i - Network accessline 19
- The prompt's `http://127.0.0.1:<port>/` refers to THIS remote host. Always use `<remote-env-url>`. - Network accessline 20
- Do NOT save remote data to files; curl and inspect in the shell only.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 21
- JSON API base: `<remote-env-url>` (read-only, **no auth**) - Health: `GET /health` -> `{"ok": true, ...}` - The prompt's `http://127.0.0.1:<port>/` and login `ops.lead@peopleops.local / PeopleOps#2026` refer to THIS remote environment. Do - Network accessline 23
- The prompt's `http://127.0.0.1:<port>/` and login `ops.lead@peopleops.local / PeopleOps#2026` refer to THIS remote environment. Do not start a local server and do not attempt login; just hit `/api/*`. - Network accessline 49
2. **Probe + confirm.** `curl -s <remote-env-url>/health` then `/api/summary` to confirm the service is live. Do not save remote data to disk; pipe through `python3 -m json.tool` or a `python3 -c` filter.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 11
## Environment contract - Web UI: `<remote-env-url>/` (Northwind People Lifecycle Portal). The prompt's `http://127.0.0.1:<port>/` and login `ops.lead@peopleops.local / PeopleOps#2026` point at THIS remote host. Credentials are illustrative - Network accessline 12
- Web UI: `<remote-env-url>/` (Northwind People Lifecycle Portal). The prompt's `http://127.0.0.1:<port>/` and login `ops.lead@peopleops.local / PeopleOps#2026` point at THIS remote host. Credentials are illustrative; no auth is needed. - Network accessline 13
- JSON API base: `<remote-env-url>` (read-only, no auth). Always use this base. - Health: `GET /health` -> `{"ok": true, ...}`. Probe it first if a call hangs. - Use `curl -s --max-time 20 "<base>/api/..."` and pipe through `python3 -m json - Network accessline 15
- Use `curl -s --max-time 20 "<base>/api/..."` and pipe through `python3 -m json.tool`. Do NOT save remote data to disk.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 11
to reconcile HR lifecycle records and return normalized business-label JSON. Signals: employee onboarding closeout, leave source precedence, payroll assignment + accrual readiness, recruitment outcome reconciliation, policy-case folder/noti - Network accessline 12
onboarding closeout, leave source precedence, payroll assignment + accrual readiness, recruitment outcome reconciliation, policy-case folder/notice review. The prompt's `http://127.0.0.1:<port>/` URL and `ops.lead@peopleops.local / PeopleOp - Network accessline 13
outcome reconciliation, policy-case folder/notice review. The prompt's `http://127.0.0.1:<port>/` - Network accessline 19
- API base: `<remote-env-url>` (read-only, no auth). Health: `GET /health`. - Never start a local service. Use `curl -s` against `/api/*`. - Network accessline 20
- Never start a local service. Use `curl -s` against `/api/*`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 17
- API base: `<remote-env-url>` (read-only, no auth). Prefixed here as `{BASE}`. - Health: `GET {BASE}/health`. - The prompt's `http://127.0.0.1:<port>/` and login creds refer to this same remote env; use - Network accessline 18
- Health: `GET {BASE}/health`. - The prompt's `http://127.0.0.1:<port>/` and login creds refer to this same remote env; use `{BASE}` for all API calls. The REST API under `/api/*` needs no auth. - Network accessline 19
- The prompt's `http://127.0.0.1:<port>/` and login creds refer to this same remote env; use
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/self/self_attempt_02/SKILL.md
- Network accessline 27
- Web UI: `<remote-env-url>/` - Data is deterministic (manifest seed 12012). GET only; never POST/PUT. - Invoke: `curl -s <remote-env-url>/api/<endpoint> | python3 -m json.tool` - Network accessline 28
- Data is deterministic (manifest seed 12012). GET only; never POST/PUT. - Invoke: `curl -s <remote-env-url>/api/<endpoint> | python3 -m json.tool` - Pipe through `python3 -c` to filter large lists (e.g. `if e['employee_id']=='EMP-122'`). - Dangerous shell operationline 29
- Invoke: `curl -s <remote-env-url>/api/<endpoint> | python3 -m json.tool` - Network accessline 29
- Invoke: `curl -s <remote-env-url>/api/<endpoint> | python3 -m json.tool`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_012/self/self_attempt_03/SKILL.md
- Network accessline 26
If the prompt mentions PeopleOps Console, login `ops.lead@peopleops.local` / `PeopleOps#2026`, an `http://127.0.0.1:<port>/` solver URL, employees - Network accessline 27
`PeopleOps#2026`, an `http://127.0.0.1:<port>/` solver URL, employees - Network accessline 39
`127.0.0.1:<port>` and credentials point at THIS remote host — substitute `<remote-env-url>` everywhere. - Always fetch with `curl -s` and pipe to `python3 -m json.tool` for readability. - Network accessline 40
`<remote-env-url>` everywhere. - Always fetch with `curl -s` and pipe to `python3 -m json.tool` for readability. - `?q=<keyword>` searches across fields (employee id, name, case id, candidate - Network accessline 41
- Always fetch with `curl -s` and pipe to `python3 -m json.tool` for readability. - Network accessline 71
field and its `allowed_values`. You will ONLY emit values from those enums — never free-text. The template IS the contract. 2. **Health check** (optional): `curl -s <remote-env-url>/health`. - Network accessline 72
never free-text. The template IS the contract. 2. **Health check** (optional): `curl -s <remote-env-url>/health`. 3. **Extract IDs** from the prompt (employee id `EMP-XXX`, case id `CASE-XXX`, - Network accessline 73
2. **Health check** (optional): `curl -s <remote-env-url>/health`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/fewshot/fewshot_attempt_02/references/portal_model.md
- Network accessline 5
Base URL example (from `environment_access.md`): `http://task-env:9013/` — replace `<TASK_ENV_BASE_URL>` with this in every request.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_013/self/self_attempt_02/SKILL.md
- Network accessline 29
- Base URL: from `environment_access.md` (e.g. `http://task-env:9013/`). No auth.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/fewshot/fewshot_attempt_03/references/environment_and_data_model.md
- Network accessline 9
From `environment_access.md` (e.g. `http://task-env:9014/`). All paths below are
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_01/SKILL.md
- Network accessline 26
- The prompt and `task_context.json` write the base URL as the placeholder `<TASK_ENV_BASE_URL>`. Substitute the real base URL from `environment_access.md` (e.g. `http://task-env:9014/`). All requests go to that host.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_01/references/endpoints.md
- Network accessline 5
The prompt and `task_context.json` carry the placeholder `<TASK_ENV_BASE_URL>`. Resolve it from `environment_access.md` in the work directory (e.g. `http://task-env:9014/`). All requests target that host. If `environment_access.md` is missi
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_01/scripts/query_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 13
Env overrides (optional): NORTHSTAR_BASE_URL e.g. http://task-env:9014 - Network accessline 14
NORTHSTAR_BASE_URL e.g. http://task-env:9014 - Network accessline 26
DEFAULT_BASE_URL = "http://task-env:9014" - Access to secretsline 32
[Credential-related source omitted] - Access to secretsline 33
[Credential-related source omitted] - Network accessline 39
url = base + SQL_PATH body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( - Network accessline 40
body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( url, - Network accessline 41
req = urllib.request.Request( - Network accessline 48
method="POST", ) with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 49
) with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 50
with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 56
if not path.startswith("/"): path = "/" + path req = urllib.request.Request(base + path, method="GET") - Network accessline 57
path = "/" + path req = urllib.request.Request(base + path, method="GET") with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 58
req = urllib.request.Request(base + path, method="GET") - Network accessline 59
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_03/SKILL.md
- Network accessline 1
--- name: northstar-payer-ops description: Northstar Health Plan payer-operations structured determinations. Use when a task asks for a JSON determination, disposition, correction packet, P2P summary, or margin-queue summary for a Northstar - Network accessline 2
name: northstar-payer-ops description: Northstar Health Plan payer-operations structured determinations. Use when a task asks for a JSON determination, disposition, correction packet, P2P summary, or margin-queue summary for a Northstar cas - Network accessline 3
description: Northstar Health Plan payer-operations structured determinations. Use when a task asks for a JSON determination, disposition, correction packet, P2P summary, or margin-queue summary for a Northstar case/appeal/claim/P2P/queue a - Network accessline 36
- **Base URL:** `http://task-env:9014/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_014/self/self_attempt_03/references/environment_and_schema.md
- Network accessline 7
- **Base URL:** `http://task-env:9014/` (from `environment_access.md`; also appears as `<TASK_ENV_BASE_URL>` in task payloads — they are the same target).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/fewshot/fewshot_attempt_01/scripts/ehr_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 71
def _get(self, path: str) -> Any: url = f"{self.base_url}/{path.lstrip('/')}" req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 72
url = f"{self.base_url}/{path.lstrip('/')}" req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 73
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 74
try: with urllib.request.urlopen(req, timeout=self.timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 75
with urllib.request.urlopen(req, timeout=self.timeout) as resp:
experiments/claude_code_glm_5_2_max/reports/skills/task_group_015/self/self_attempt_03/scripts/fetch_patient_bundle.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 5
# # Usage: fetch_patient_bundle.sh <base_url> <patient_id> # Example: fetch_patient_bundle.sh "http://task-env:9015" "P-12345" - Network accessline 6
# Usage: fetch_patient_bundle.sh <base_url> <patient_id> # Example: fetch_patient_bundle.sh "http://task-env:9015" "P-12345" # - Network accessline 7
# Example: fetch_patient_bundle.sh "http://task-env:9015" "P-12345" - Network accessline 23
if [ -z "$sub" ]; then echo "=== patient $PID ===" curl -s -m 20 "${BASE}/api/patients/${PID}" - Network accessline 24
echo "=== patient $PID ===" curl -s -m 20 "${BASE}/api/patients/${PID}" else - Network accessline 25
curl -s -m 20 "${BASE}/api/patients/${PID}" - Network accessline 26
else echo "=== /api/patients/${PID}/${sub} ===" curl -s -m 20 "${BASE}/api/patients/${PID}/${sub}" - Network accessline 27
echo "=== /api/patients/${PID}/${sub} ===" curl -s -m 20 "${BASE}/api/patients/${PID}/${sub}" fi - Network accessline 28
curl -s -m 20 "${BASE}/api/patients/${PID}/${sub}"
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/fewshot/fewshot_attempt_01/environment_contract.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/self/self_attempt_01/references/runtime_endpoints.md
- Network accessline 5
## Base URL `http://task-env:9016/` — referred to in prompts as `<TASK_ENV_BASE_URL>`. Substitute the base URL from the current run's access file. - Network accessline 6
`http://task-env:9016/` — referred to in prompts as `<TASK_ENV_BASE_URL>`. Substitute the base URL from the current run's access file.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_016/self/self_attempt_02/references/runtime_endpoints.md
- Network accessline 5
## Base URL Provided in `environment_access.md` (training run: `http://task-env:9016/`). Substitute it wherever a prompt writes `<TASK_ENV_BASE_URL>`. Do not hardcode it in any response or skill output. - Network accessline 6
Provided in `environment_access.md` (training run: `http://task-env:9016/`). Substitute it wherever a prompt writes `<TASK_ENV_BASE_URL>`. Do not hardcode it in any response or skill output.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/fewshot/fewshot_attempt_03/references/hub_endpoints.md
- Network accessline 9
- Base URL: given in `environment_access.md` (e.g. `http://task-env:9017/`).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02/reference/validate.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 1
--- name: investigation-review-hub-gap-analysis description: Produce a schema-conformant JSON gap-analysis / remediation dashboard from the Investigation Review Hub for a legal-discovery matter. Use when a task points at the Investigation R - Network accessline 2
name: investigation-review-hub-gap-analysis description: Produce a schema-conformant JSON gap-analysis / remediation dashboard from the Investigation Review Hub for a legal-discovery matter. Use when a task points at the Investigation Revie - Network accessline 3
description: Produce a schema-conformant JSON gap-analysis / remediation dashboard from the Investigation Review Hub for a legal-discovery matter. Use when a task points at the Investigation Review Hub (base URL like http://task-env:PORT/)
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_01/references/hub_schema.md
- Network accessline 9
- Base URL: `http://task-env:9017/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_02/SKILL.md
- Network accessline 1
--- name: investigation-review-hub-analysis description: Produce a structured-JSON gap / retention / remediation / production-readiness analysis for a legal-discovery matter by querying the shared Investigation Review Hub. Use whenever a ta - Network accessline 2
name: investigation-review-hub-analysis description: Produce a structured-JSON gap / retention / remediation / production-readiness analysis for a legal-discovery matter by querying the shared Investigation Review Hub. Use whenever a task p - Network accessline 3
description: Produce a structured-JSON gap / retention / remediation / production-readiness analysis for a legal-discovery matter by querying the shared Investigation Review Hub. Use whenever a task points at the Review Hub at http://task-e - Network accessline 8
This skill turns a matter-level legal-discovery review request into a single JSON object that conforms exactly to the task's `answer_template.json`. The Review Hub at `http://task-env:9017/` is the **only** source of business evidence. The
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_02/reference/hub_endpoints.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_03/SKILL.md
- Network accessline 1
--- name: investigation-review-hub-answer description: Produce a single-JSON-object answer for an e-discovery / regulatory investigation review task served by the Investigation Review Hub (http://task-env:9017/). Use when a task asks for a - Network accessline 2
name: investigation-review-hub-answer description: Produce a single-JSON-object answer for an e-discovery / regulatory investigation review task served by the Investigation Review Hub (http://task-env:9017/). Use when a task asks for a prod - Network accessline 3
description: Produce a single-JSON-object answer for an e-discovery / regulatory investigation review task served by the Investigation Review Hub (http://task-env:9017/). Use when a task asks for a production-readiness, retention/preservati - Network accessline 15
- A **context/scope payload** (`request_context.json`, `review_scope.json`, or `matter_context.json`) carrying the matter ID, client, category labels/family, and the base URL. - An **`answer_template.json`** that is the *authoritative* outp - Network accessline 17
- The **base URL** `http://task-env:9017/`. - Network accessline 28
Base URL: `http://task-env:9017/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_017/self/self_attempt_03/references/hub_endpoints.md
- Network accessline 3
Base URL: `http://task-env:9017/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/fewshot/fewshot_attempt_01/references/portal_endpoints.md
- Network accessline 3
The Court Operations Portal is the network source of truth. Reach it only via the base URL and allowed endpoint list in `environment_access.md` (default base: `http://task-env:9018/`). It is open within the network — no credentials. Use `GE
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/fewshot/fewshot_attempt_03/references/portal_endpoints.md
- Network accessline 7
`http://task-env:9018/` (read from `environment_access.md`; the host may differ per environment — always read the file rather than assuming.)
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 22
Base URL: `http://task-env:9018/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/reflect-3/reflect-3_attempt_03/portal_endpoints.md
- Network accessline 3
Base URL: `http://task-env:9018/`
experiments/claude_code_glm_5_2_max/reports/skills/task_group_018/self/self_attempt_01/SKILL.md
- Network accessline 20
`environment_access.md` authorizes a base URL (e.g. `http://task-env:9018/`) and a fixed allow-list of `GET /api/...` endpoints. Credentials are none / open within network.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/fewshot/fewshot_attempt_03/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/self/self_attempt_01/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_019/self/self_attempt_03/SKILL.md
- Network accessline 9
- **Base URL**: Use the `<TASK_ENV_BASE_URL>` placeholder from the prompt (resolved at task time; typically `http://task-env:9019/`).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/fewshot/fewshot_attempt_03/SKILL.md
- Access to secretsline 40
[Credential-related source omitted]
experiments/claude_code_glm_5_2_max/reports/skills/task_group_020/fewshot/fewshot_attempt_03/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 8
python3 fetch_deal.py --deal-id PRJ_EXAMPLE python3 fetch_deal.py --deal-id PRJ_EXAMPLE --env-file /work/environment_access.md python3 fetch_deal.py --deal-id PRJ_EXAMPLE --base-url http://host:port/ --token X - Network accessline 10
python3 fetch_deal.py --deal-id PRJ_EXAMPLE --base-url http://host:port/ --token X - Network accessline 41
def http_get_json(url, timeout=30): req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 43
with urllib.request.urlopen(req, timeout=timeout) as resp:
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/fewshot/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01/reference/canon.py
- Executable scriptline 1
Executable file
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01/reference/hub_client.py
- Executable scriptline 1
Executable file - Access to secretsline 18
[Credential-related source omitted] - Access to secretsline 19
[Credential-related source omitted] - Network accessline 56
[Credential-related source omitted] - Network accessline 57
[Credential-related source omitted] - Network accessline 58
[Credential-related source omitted] - Network accessline 59
with urllib.request.urlopen(req) as r: - Network accessline 64
if not _BASE: raise RuntimeError("call configure_from_file(env_access_path) first") req = urllib.request.Request( - Network accessline 65
raise RuntimeError("call configure_from_file(env_access_path) first") req = urllib.request.Request( _BASE + path, - Network accessline 66
req = urllib.request.Request( - Network accessline 71
) try: with urllib.request.urlopen(req) as r: - Network accessline 72
try: with urllib.request.urlopen(req) as r: return json.loads(r.read().decode()) - Network accessline 73
with urllib.request.urlopen(req) as r:
experiments/claude_code_glm_5_2_max/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03/reference/api_query_guide.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 73
## References (load as needed) - `references/environment_and_sql.md` — endpoint mechanics, curl patterns, response shapes, transaction constraints. - Network accessline 74
- `references/environment_and_sql.md` — endpoint mechanics, curl patterns, response shapes, transaction constraints.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_01/references/environment_and_sql.md
- Network accessline 3
This file describes how to reach the running Atlas Commerce Operations service. The host, token, and exact endpoint rules live in `environment_access.md` at the workspace root — **read it fresh each task**; do not rely on the values below, - Network accessline 14
[Credential-related source omitted] - Network accessline 21
[Credential-related source omitted] - Network accessline 27
[Credential-related source omitted] - Network accessline 32
curl -sS -m 60 -X POST "$GDPEVO_ENV_BASE_URL/api/sql" \ - Network accessline 44
curl -sS -m 60 -X POST "$GDPEVO_ENV_BASE_URL/api/sql/transaction" \
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 110
- `environment_api.md` — full endpoint contract, request/response shapes, and worked curl forms.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_02/environment_api.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 25
[Credential-related source omitted]
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/self/self_attempt_01/references/api_usage.md
- Network accessline 8
# Parse environment_access.md for the base URL and token the env provides. export GDPEVO_ENV_BASE_URL="http://task-env:9022/" # value from environment_access.md - Network accessline 9
export GDPEVO_ENV_BASE_URL="http://task-env:9022/" # value from environment_access.md - Network accessline 19
curl -sS "$GDPEVO_ENV_BASE_URL/api/schema" -H "$AUTH" - Network accessline 27
curl -sS "$GDPEVO_ENV_BASE_URL/api/data-dictionary" -H "$AUTH" - Network accessline 35
curl -sS "$GDPEVO_ENV_BASE_URL/api/correction-audit" -H "$AUTH" - Network accessline 48
curl -sS -X POST "$GDPEVO_ENV_BASE_URL/api/sql" -H "$AUTH" -H "$JSON_CT" \ - Network accessline 71
curl -sS -X POST "$GDPEVO_ENV_BASE_URL/api/sql/transaction" -H "$AUTH" -H "$JSON_CT" \
experiments/claude_code_glm_5_2_max/reports/skills/task_group_022/self/self_attempt_02/reference/endpoints.md
- Network accessline 13
The literal base URL is printed in `environment_access.md` (e.g. `http://task-env:9022/`). Prefer the `<TASK_ENV_BASE_URL>` / `GDPEVO_ENV_BASE_URL` env var when it is populated; if that env var is empty, fall back to the literal URL from th - Network accessline 32
[Credential-related source omitted] - Network accessline 34
[Credential-related source omitted] - Network accessline 58
curl -sS -X POST "$BASE/api/sql/transaction" \ - Network accessline 65
- Always pass `curl -sS` and pipe through `python3 -m json.tool` (or `jq`) to read responses reliably.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01/references/module_methods.md
- Dangerous shell operationline 324
magnitude = SE_b · sqrt(df · rY · rM / (1 − rM))
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_01/references/portal.md
- Network accessline 3
The Public Health Observatory (PHO) Data Portal is the **only authorized evidence source**. Reach it only through `environment_access.md`: the base URL (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`) and the allow-list of - Network accessline 4
source**. Reach it only through `environment_access.md`: the base URL (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`) and the allow-list of endpoints. No other network source is permitted. - Network accessline 5
(`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`) and the allow-list of
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_03/references/methods.md
- Dangerous shell operationline 248
`magnitude = SE_b * sqrt(df * rY * rM / (1 - rM))`.
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/fewshot/fewshot_attempt_03/references/portal.md
- Network accessline 10
- `environment_access.md` contains `GDPEVO_ENV_BASE_URL=http://task-env:9023/` (or equivalent).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03/SKILL.md
experiments/claude_code_glm_5_2_max/reports/skills/task_group_023/self/self_attempt_02/references/portal_and_evidence.md
- Network accessline 7
- Resolve the base URL from `environment_access.md`: `<TASK_ENV_BASE_URL>` maps to `GDPEVO_ENV_BASE_URL` (e.g. `http://task-env:9023/`).
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01/references/environment_and_endpoints.md
- Network accessline 3
## Reading `environment_access.md` `environment_access.md` contains, in order: - `GDPEVO_ENV_BASE_URL=...` — the base URL (e.g. `http://task-env:9024/`). Prefix every endpoint - Network accessline 4
`environment_access.md` contains, in order: - `GDPEVO_ENV_BASE_URL=...` — the base URL (e.g. `http://task-env:9024/`). Prefix every endpoint with it. The prompt writes this as `<TASK_ENV_BASE_URL>`; substitute the value from this file. - Network accessline 5
- `GDPEVO_ENV_BASE_URL=...` — the base URL (e.g. `http://task-env:9024/`). Prefix every endpoint - Network accessline 96
[Credential-related source omitted] - Network accessline 97
[Credential-related source omitted] - Network accessline 98
[Credential-related source omitted] - Network accessline 99
[Credential-related source omitted]
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/fewshot/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/reflect-3/reflect-3_attempt_03/classify.py
- Executable scriptline 1
Executable file
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/self/self_attempt_01/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_glm_5_2_max/reports/skills/task_group_024/self/self_attempt_03/references/access_and_query.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 5
- **Base URL**: `http://34.46.77.124:8001` (from `environment_access.md`; do not use localhost).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 9
- Use the base URL supplied by the task runner (e.g. `http://34.46.77.124:8001`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 9
- Base URL is provided by the runner or environment file (e.g., `http://34.46.77.124:8001`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 4
- Read `MANIFEST.md`, `environment_access.md`, and the staged train prompt/answer template. - **Do not assume train tasks share the same schema.** Each train task may use a different event/show/batch and a different JSON template. - The rem - Network accessline 6
- The remote base URL is `http://34.46.77.124:8001`. Do not use localhost unless explicitly overridden.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_001/self/self_attempt_01/SKILL.md
- Network accessline 13
- **Base URL**: Use only `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8001`. Do NOT use localhost unless the remote URL itself points there.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 9
- Read `input/prompt.txt` for task instructions and entity IDs. - Read `input/payloads/answer_template.json` for the exact output schema and enum values. - The API base URL is typically `http://127.0.0.1:${PORT}` or similar; the prompt will - Network accessline 11
- The API base URL is typically `http://127.0.0.1:${PORT}` or similar; the prompt will specify the host and port. Treat the date given in the prompt as the current business date. - Network accessline 125
1. **Hard-coding output shape** — Always read `answer_template.json`; templates vary even within the same task family (e.g., train_003 vs. train_005). 2. **Ignoring the API port** — The prompt often references a runner-provided port or `htt - Network accessline 126
2. **Ignoring the API port** — The prompt often references a runner-provided port or `http://127.0.0.1:${PORT}`; substitute the actual value.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 11
Base URL is provided by the task runner (e.g. `http://127.0.0.1:${PORT}` or `http://34.46.77.124:8002`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/self/self_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_002/self/self_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/fewshot/fewshot_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 3
## 1. Environment & Endpoints - **CRM Service API base URL** is provided in `environment_access.md` (e.g., `http://host.docker.internal:8123` in Docker contexts). - Network accessline 4
- **CRM Service API base URL** is provided in `environment_access.md` (e.g., `http://host.docker.internal:8123` in Docker contexts).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
- Network accessline 3
## 1. Environment & Base URL - The remote support-console API is reachable **only** at the URL provided in `environment_access.md` (e.g. `http://34.46.77.124:8003`). - Network accessline 4
- The remote support-console API is reachable **only** at the URL provided in `environment_access.md` (e.g. `http://34.46.77.124:8003`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_003/self/self_attempt_03/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 5
This skill covers tasks that consume the **ApexCloud Retention Operations API** (`http://127.0.0.1:8074`) to produce structured CRM analytics deliverables for Customer Success, Finance, and Data Science teams. The API returns account profil - Network accessline 13
- **Base URL:** `http://127.0.0.1:8074`
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_004/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/fewshot/fewshot_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/fewshot/fewshot_attempt_03/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 5
- Read `input/payloads/answer_template.json` to understand the exact output schema. - Read any additional local payload files (e.g., `.json`, `.csv`) that provide batch context or reference data. - The environment base URL is `http://34.46. - Network accessline 7
- The environment base URL is `http://34.46.77.124:8005`. Use this as the source of record for all ERP/compliance data.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_005/self/self_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 5
Tasks in this group require querying a ProcureOps GraphQL API (`https://graphql-procureops.internal/v1`) and producing structured JSON outputs that follow a provided answer template. The domain spans procurement programs, supplier nominatio - Network accessline 9
- **Endpoint**: `https://graphql-procureops.internal/v1` - **Auth**: Header `x-internal-auth: test-token` - **Tool**: Use `curl` with `-s -S` and `-H` flags for the auth header and `Content-Type: application/json`
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/fewshot/fewshot_attempt_03/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 8
### Base URL Always use the remote API at `http://34.46.77.124:8006` (or the URL specified in `environment_access.md`). Never use localhost unless explicitly directed. - Network accessline 9
Always use the remote API at `http://34.46.77.124:8006` (or the URL specified in `environment_access.md`). Never use localhost unless explicitly directed. - Dangerous shell operationline 28
curl -s "http://34.46.77.124:8006/purchase_orders" | python3 -m json.tool - Network accessline 28
curl -s "http://34.46.77.124:8006/purchase_orders" | python3 -m json.tool - Network accessline 33
curl -s "http://34.46.77.124:8006/purchase_orders?po_id=PO-AX17-4481" - Network accessline 34
curl -s "http://34.46.77.124:8006/ap_invoices?po_id=PO-AX17-4481" - Network accessline 35
curl -s "http://34.46.77.124:8006/receipts?po_id=PO-AX17-4481"
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
- Network accessline 5
- **Always use** the API base URL from `environment_access.md` (e.g., `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8006`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_02/SKILL.md
- Network accessline 9
- **Base URL**: Provided by the task runner, or start the local environment at `http://127.0.0.1:8006`.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_006/self/self_attempt_03/SKILL.md
- Network accessline 19
http://127.0.0.1:8006
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 3
## Overview Tasks in this group involve querying a shared Northwind ERP API (`http://127.0.0.1:8007`) and producing strictly-structured JSON outputs. There are four main task families: - Network accessline 4
Tasks in this group involve querying a shared Northwind ERP API (`http://127.0.0.1:8007`) and producing strictly-structured JSON outputs. There are four main task families: - Network accessline 10
## Environment & API Setup - Start the API with `bash setup.sh start` or `python server.py --host 127.0.0.1 --port 8007`. - Base URL: `http://127.0.0.1:8007` - Network accessline 11
- Start the API with `bash setup.sh start` or `python server.py --host 127.0.0.1 --port 8007`. - Base URL: `http://127.0.0.1:8007` - Query live endpoints; never rely on cached snapshots. Relevant endpoints include `/orders`, `/products`, `/ - Network accessline 12
- Base URL: `http://127.0.0.1:8007`
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 3
## Overview These tasks require querying a shared Northwind ERP API (default `http://127.0.0.1:8007`) and producing strictly-formatted JSON decisions. The output schema is always defined in `input/payloads/answer_template.json` and must be - Network accessline 4
These tasks require querying a shared Northwind ERP API (default `http://127.0.0.1:8007`) and producing strictly-formatted JSON decisions. The output schema is always defined in `input/payloads/answer_template.json` and must be matched exac - Network accessline 8
2. Read the answer template (`input/payloads/answer_template.json`) to learn the exact required keys, types, enums, and sorting rules. 3. Read any memo/request payload in `input/payloads/` for target IDs, date windows, or policy rules. 4. S - Network accessline 9
3. Read any memo/request payload in `input/payloads/` for target IDs, date windows, or policy rules. 4. Start or connect to the ERP API (base URL from prompt/memo, typically `http://127.0.0.1:8007`). 5. Fetch live records for every entity r - Network accessline 10
4. Start or connect to the ERP API (base URL from prompt/memo, typically `http://127.0.0.1:8007`). 5. Fetch live records for every entity referenced. Do not use cached or assumed inventory snapshots. 6. Apply the business rules described in - Network accessline 16
## API Usage Habits - **Base URL**: Usually `http://127.0.0.1:8007`; may be provided in the memo. Start the environment from `task_group/task_group_007/env` if needed. - Network accessline 17
- **Base URL**: Usually `http://127.0.0.1:8007`; may be provided in the memo. Start the environment from `task_group/task_group_007/env` if needed.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 3
## 1. Environment & API Basics - Start the shared ERP API from `task_group/task_group_007/env` with `bash setup.sh start`, or run `python server.py --host 127.0.0.1 --port 8007`. - Base URL: `http://127.0.0.1:8007`. - Network accessline 4
- Start the shared ERP API from `task_group/task_group_007/env` with `bash setup.sh start`, or run `python server.py --host 127.0.0.1 --port 8007`. - Base URL: `http://127.0.0.1:8007`. - Use only the public API endpoints; do not inspect env - Network accessline 5
- Base URL: `http://127.0.0.1:8007`.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/reflect-3/reflect-3_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 3
## API Access Habits - **Base URL**: Use the remote ERP URL from `environment_access.md` (e.g. `http://34.46.77.124:8007`). Do not default to `localhost` or `127.0.0.1` unless the access file explicitly says so. - Network accessline 4
- **Base URL**: Use the remote ERP URL from `environment_access.md` (e.g. `http://34.46.77.124:8007`). Do not default to `localhost` or `127.0.0.1` unless the access file explicitly says so.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_007/self/self_attempt_02/SKILL.md
- Network accessline 5
- **Read `environment_access.md` first.** It overrides any localhost/127.0.0.1 references in task prompts. Use only the remote base URL declared there (e.g. `http://34.46.77.124:8007`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_009/self/self_attempt_03/SKILL.md
- Network accessline 3
## Remote API Entrypoint - **Always use** `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8009` as the authoritative base URL. - Network accessline 4
- **Always use** `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8009` as the authoritative base URL. - Network accessline 5
- **Override rule:** If `payloads/environment_access.json` lists `http://127.0.0.1:8047` or any localhost reference, ignore it and use the remote URL above.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/reflect-3/reflect-3_attempt_02/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/reflect-3/reflect-3_attempt_03/SKILL.md
- Network accessline 9
1. **Base URL**: Use only the provided environment base URL (e.g., `http://34.46.77.124:8010`). Do not use localhost or 127.0.0.1.
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/self/self_attempt_01/SKILL.md
- Network accessline 6
## Prerequisites - The task provides `input/prompt.txt`, `input/payloads/<request>.json`, and `input/payloads/answer_template.json` - Environment URL is fixed: `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8010` (or as specified in `environment_ - Network accessline 7
- The task provides `input/prompt.txt`, `input/payloads/<request>.json`, and `input/payloads/answer_template.json` - Environment URL is fixed: `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8010` (or as specified in `environment_access.md`) - Do - Network accessline 8
- Environment URL is fixed: `GDPEVO_ENV_BASE_URL=http://34.46.77.124:8010` (or as specified in `environment_access.md`) - Network accessline 25
curl -s http://34.46.77.124:8010/api/catalog - Network accessline 31
curl -s http://34.46.77.124:8010/api/portfolios/<portfolio_id> - Network accessline 37
curl -s http://34.46.77.124:8010/api/policies - Network accessline 43
curl -s http://34.46.77.124:8010/api/instruments/bonds - Network accessline 49
curl -s http://34.46.77.124:8010/api/issuers - Network accessline 55
curl -s http://34.46.77.124:8010/api/market/energy - Network accessline 61
curl -s http://34.46.77.124:8010/api/indices - Network accessline 62
curl -s http://34.46.77.124:8010/api/index-levels/<index_id> - Network accessline 68
curl -s http://34.46.77.124:8010/api/allocation/opportunity-sets - Network accessline 69
curl -s http://34.46.77.124:8010/api/allocation/prior-views - Network accessline 70
curl -s http://34.46.77.124:8010/api/macro-signals - Network accessline 146
BASE="http://34.46.77.124:8010" - Network accessline 149
# 1. Catalog curl -s "$BASE/api/catalog" | jq . - Network accessline 150
curl -s "$BASE/api/catalog" | jq . - Network accessline 152
# 2. Portfolio holdings & constraints curl -s "$BASE/api/portfolios/$PORTFOLIO" | jq . - Network accessline 153
curl -s "$BASE/api/portfolios/$PORTFOLIO" | jq . - Network accessline 155
# 3. Bond universe for candidate selection curl -s "$BASE/api/instruments/bonds" | jq '.[] | select(.candidate==true)' - Network accessline 156
curl -s "$BASE/api/instruments/bonds" | jq '.[] | select(.candidate==true)' - Network accessline 158
# 4. Issuer watchlist status curl -s "$BASE/api/issuers" | jq '.[] | select(.watchlist==true)' - Network accessline 159
curl -s "$BASE/api/issuers" | jq '.[] | select(.watchlist==true)' - Network accessline 161
# 5. Index levels for correlation curl -s "$BASE/api/index-levels/IDX_EM" | jq '.levels' - Network accessline 162
curl -s "$BASE/api/index-levels/IDX_EM" | jq '.levels'
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_010/self/self_attempt_02/SKILL.md
- Network accessline 211
# 2. Explore API BASE="http://<host>:<port>" curl -s "$BASE/" | grep -oE '/api/[^<]+' - Network accessline 213
curl -s "$BASE/" | grep -oE '/api/[^<]+' - Network accessline 215
# 3. Fetch data curl -s "$BASE/api/portfolios/PF-EXAMPLE" > portfolio.json - Network accessline 216
curl -s "$BASE/api/portfolios/PF-EXAMPLE" > portfolio.json - Network accessline 217
curl -s "$BASE/api/policies" > policies.json - Network accessline 218
curl -s "$BASE/api/index-levels/IDX_EXAMPLE" > levels.json - Network accessline 219
curl -s "$BASE/api/instruments/bonds" > bonds.json - Network accessline 220
curl -s "$BASE/api/macro-signals" > signals.json
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 7
- Base URL is given by the task prompt or `GDPEVO_ENV_BASE_URL` (e.g. `http://34.46.77.124:8011`).
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_011/self/self_attempt_03/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/fewshot/fewshot_attempt_03/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_kimi2_6_enabled/reports/skills/task_group_012/self/self_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 40
Base URL is supplied by the runner (e.g. `http://127.0.0.1:8080` or `:8067`); read it from the task's environment/prompt. All responses are JSON. Use `curl`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 36
Base URL is supplied by the runner (e.g. `http://127.0.0.1:8080` or `:8067`). Use `curl -s`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 232
Base URL is in `environment_access.md` (e.g. `http://127.0.0.1:8080`; some prompts cite a different port — use the runner-supplied base). All responses are JSON. Use `curl -s`. Start with: - Network accessline 233
port — use the runner-supplied base). All responses are JSON. Use `curl -s`. Start with: - Network accessline 236
curl -s $BASE/health # sanity: record counts - Network accessline 237
curl -s $BASE/api/policies # controlled enums in force
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 12
## Data API (HTTP only — use curl or python urllib; never WebFetch/HTTPS)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_001/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 11
## Environment / API (HTTP-only; use curl or python urllib, never WebFetch)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 5
(CRM, quotes, freight/logistics, milestone receivables). Use this skill WHENEVER a task asks you to verify customer/quote/RFQ/opportunity records against a MedBridge Sales Ops API (base like http://127.0.0.1:<PORT>) and emit JSON matchi - Network accessline 6
task asks you to verify customer/quote/RFQ/opportunity records against a MedBridge Sales Ops API (base like http://127.0.0.1:<PORT>) and emit JSON matching an answer_template.json — including: catalog-tier quote pricing and EXW totals; - Network accessline 7
Sales Ops API (base like http://127.0.0.1:<PORT>) and emit JSON matching an - Network accessline 34
2. **API only.** Never read local business files. Reach the API with `curl`. The base URL is given in the prompt (env var like `API_BASE_URL` / `BASE_URL` or `http://127.0.0.1:<PORT>`). Confirm reachability with `GET /health`. - Network accessline 35
base URL is given in the prompt (env var like `API_BASE_URL` / `BASE_URL` or `http://127.0.0.1:<PORT>`). Confirm reachability with `GET /health`. 3. **Fetch full collections, then filter yourself.** List filters - Network accessline 36
`http://127.0.0.1:<PORT>`). Confirm reachability with `GET /health`. 3. **Fetch full collections, then filter yourself.** List filters (`?quote_id=...`, `?customer_id=...`) match a value **anywhere/nested** in the
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 9
opportunity/invoice/payment/revenue-recognition reconciliation with CRM follow-up routing — especially when the prompt references a MedBridge / "Sales Ops" API base URL (e.g. API_BASE_URL / BASE_URL / http://127.0.0.1:<PORT>), - Network accessline 10
follow-up routing — especially when the prompt references a MedBridge / "Sales Ops" API base URL (e.g. API_BASE_URL / BASE_URL / http://127.0.0.1:<PORT>), quote IDs (Q-...), RFQ IDs (RFQ-...), freight IDs (FR-...), opportunity IDs - Network accessline 11
Ops" API base URL (e.g. API_BASE_URL / BASE_URL / http://127.0.0.1:<PORT>), - Network accessline 54
Base URL comes from the runner (env `API_BASE_URL` / `BASE_URL`, or `http://127.0.0.1:<PORT>`). Reach it with `curl -s`. - Network accessline 55
`http://127.0.0.1:<PORT>`). Reach it with `curl -s`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 18
## The data API - Base URL is provided by the runner (env var named e.g. `API_BASE_URL` / `BASE_URL` / `http://127.0.0.1:<PORT>`). Read-only JSON over HTTP (use curl or python urllib). - Network accessline 19
- Base URL is provided by the runner (env var named e.g. `API_BASE_URL` / `BASE_URL` / `http://127.0.0.1:<PORT>`). Read-only JSON over HTTP (use curl or python urllib). - Collections (GET): `/api/customers`, `/api/products`, `/api/rfqs`, - Network accessline 20
`http://127.0.0.1:<PORT>`). Read-only JSON over HTTP (use curl or python urllib).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/reflect-3/reflect-3_attempt_03/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/self/self_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_002/self/self_attempt_03/SKILL.md
- Network accessline 14
## The remote API - Base URL: `<remote-env-url>` (or the env var the runner provides: `API_BASE_URL` / `BASE_URL` / `http://127.0.0.1:${PORT}`). Read-only JSON over HTTP; - Network accessline 15
- Base URL: `<remote-env-url>` (or the env var the runner provides: `API_BASE_URL` / `BASE_URL` / `http://127.0.0.1:${PORT}`). Read-only JSON over HTTP; use `curl -s` or python urllib. There are NO local data files. - Network accessline 16
`API_BASE_URL` / `BASE_URL` / `http://127.0.0.1:${PORT}`). Read-only JSON over HTTP; use `curl -s` or python urllib. There are NO local data files. - Collections: customers, products, rfqs, quotes, freight-quotes, policies, - Network accessline 17
use `curl -s` or python urllib. There are NO local data files.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 33
2. Confirm the API base URL. The harness may override the prompt's URL via an `environment_access.md` file or instructions; **that override wins**. Train prompts said `8057` but the real base was `http://127.0.0.1:8086`. If unsure, - Network accessline 34
`environment_access.md` file or instructions; **that override wins**. Train prompts said `8057` but the real base was `http://127.0.0.1:8086`. If unsure, `curl <base>/health` and `curl <base>/api/catalog` to confirm it is live and - Network accessline 35
prompts said `8057` but the real base was `http://127.0.0.1:8086`. If unsure, `curl <base>/health` and `curl <base>/api/catalog` to confirm it is live and to see available endpoints/record counts. - Network accessline 36
`curl <base>/health` and `curl <base>/api/catalog` to confirm it is live and
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 53
- Base URL is **`http://127.0.0.1:8086`** for ALL lookups. It overrides any
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 33
only those keys, only those enum spellings, and respect its stated ordering ("preserve payload order" vs "ascending case_id order"). 2. **Fix the base URL.** Prompts often say `http://127.0.0.1:8057`. The harness - Network accessline 34
("preserve payload order" vs "ascending case_id order"). 2. **Fix the base URL.** Prompts often say `http://127.0.0.1:8057`. The harness overrides this — use the base URL from `environment_access.md` (currently - Network accessline 35
2. **Fix the base URL.** Prompts often say `http://127.0.0.1:8057`. The harness - Network accessline 36
overrides this — use the base URL from `environment_access.md` (currently `http://127.0.0.1:8086`). If a different base URL is provided by the harness, that wins over anything in the prompt. - Network accessline 37
`http://127.0.0.1:8086`). If a different base URL is provided by the harness, - Network accessline 40
GET its record plus all dependent records (account, outage, diagnostics, troubleshooting, line, device, bill, plan, SLA, export-runs, messages) BEFORE deciding. Use `curl -s <base>/api/...`. Endpoints are listed in - Network accessline 41
troubleshooting, line, device, bill, plan, SLA, export-runs, messages) BEFORE deciding. Use `curl -s <base>/api/...`. Endpoints are listed in `environment_access.md` and `/api/catalog`. `/api/search?q=` and - Network accessline 42
deciding. Use `curl -s <base>/api/...`. Endpoints are listed in
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_01/SKILL.md
- Network accessline 12
## API base URL & lookup habits - Base URL: `<remote-env-url>` . **Always use this**, even when the prompt text says `http://127.0.0.1:8057` (the prompt's URL is a placeholder). - Network accessline 13
- Base URL: `<remote-env-url>` . **Always use this**, even when the prompt text says `http://127.0.0.1:8057` (the prompt's URL is a placeholder). - It is read-only. Confirm with `curl -s <base>/health`. `GET /api/catalog` lists - Network accessline 14
prompt text says `http://127.0.0.1:8057` (the prompt's URL is a placeholder). - It is read-only. Confirm with `curl -s <base>/health`. `GET /api/catalog` lists endpoints + record counts. - Network accessline 15
- It is read-only. Confirm with `curl -s <base>/health`. `GET /api/catalog` lists
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_02/SKILL.md
- Network accessline 12
- **Base URL: `<remote-env-url>`** (overrides any `127.0.0.1:8057` shown in prompts). Read-only. Confirm with `GET /health`, list with `GET /api/catalog`. - Call with curl, e.g. `curl -s <remote-env-url>api/tickets/TCK-6101`. - Network accessline 13
shown in prompts). Read-only. Confirm with `GET /health`, list with `GET /api/catalog`. - Call with curl, e.g. `curl -s <remote-env-url>api/tickets/TCK-6101`. - A missing record returns `{"error":"not_found"}` (HTTP 200) — treat that as the - Network accessline 14
- Call with curl, e.g. `curl -s <remote-env-url>api/tickets/TCK-6101`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_003/self/self_attempt_03/SKILL.md
- Network accessline 13
## API usage habits Base URL is given in `environment_access.md` and **overrides** any `http://127.0.0.1:8057` in the prompt. Call with `curl -s <base>/<endpoint>`. The API is read-only; it never exposes answers. Confirm reachability with ` - Network accessline 14
Base URL is given in `environment_access.md` and **overrides** any `http://127.0.0.1:8057` in the prompt. Call with `curl -s <base>/<endpoint>`. The API is read-only; it never exposes answers. Confirm reachability with `GET /health` and use
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 3
description: >- Standard operating procedure for the ApexCloud Retention Operations API (http://127.0.0.1:8074). Use this whenever a task asks you to build a retention - Network accessline 4
Standard operating procedure for the ApexCloud Retention Operations API (http://127.0.0.1:8074). Use this whenever a task asks you to build a retention artifact for ApexCloud accounts: renewal/churn risk queues, QBR metric packets, - Network accessline 5
(http://127.0.0.1:8074). Use this whenever a task asks you to build a retention - Network accessline 30
- **Base URL:** `http://127.0.0.1:8074`. It is already running; never start it, and ignore any `setup.sh` / file-path instructions inside a task prompt. - **Read-only HTTP only** (e.g. `curl`). Do not look for local data files or `env/`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_01/scripts/apex_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 20
BASE = "http://127.0.0.1:8074" - Network accessline 24
def _get(path): with urllib.request.urlopen(BASE + path, timeout=30) as r: - Network accessline 25
with urllib.request.urlopen(BASE + path, timeout=30) as r: - Network accessline 29
def _get_csv(path): with urllib.request.urlopen(BASE + path, timeout=30) as r: - Network accessline 30
with urllib.request.urlopen(BASE + path, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 2
name: apexcloud-retention-ops description: >- Operating procedure for the ApexCloud Retention Operations API (http://127.0.0.1:8074). Use this - Network accessline 3
description: >- Operating procedure for the ApexCloud Retention Operations API (http://127.0.0.1:8074). Use this skill WHENEVER a task asks you to build, score, or summarize anything in the ApexCloud retention / - Network accessline 4
Operating procedure for the ApexCloud Retention Operations API (http://127.0.0.1:8074). Use this - Network accessline 25
GET http://127.0.0.1:8074/api/health # row counts + seed, good sanity check
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_02/references/playbook.md
- Network accessline 20
Base URL `http://127.0.0.1:8074`. All GET, JSON unless noted. The service is always running.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 3
description: >- Operating procedure for solving ApexCloud Retention Operations tasks against the read-only HTTP API at http://127.0.0.1:8074. Use this skill whenever a task asks you - Network accessline 4
Operating procedure for solving ApexCloud Retention Operations tasks against the read-only HTTP API at http://127.0.0.1:8074. Use this skill whenever a task asks you to build a renewal risk queue, a retention action board, a QBR metrics - Network accessline 5
read-only HTTP API at http://127.0.0.1:8074. Use this skill whenever a task asks you - Network accessline 23
All data comes from the read-only HTTP API. The base URL is given in the prompt / environment (typically `http://127.0.0.1:8074`). Use `curl` + a JSON parser - Network accessline 24
environment (typically `http://127.0.0.1:8074`). Use `curl` + a JSON parser
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_03/references/api_and_rules.md
- Network accessline 3
Read this when you need the exact shape of a response, to confirm a threshold, or to copy a worked calculation. Base URL is in the prompt (typically `http://127.0.0.1:8074`). All endpoints are HTTP GET, JSON unless noted. Read-only. - Network accessline 5
`http://127.0.0.1:8074`). All endpoints are HTTP GET, JSON unless noted. Read-only. - Network accessline 205
BASE=http://127.0.0.1:8074 curl -s "$BASE/api/health" - Network accessline 206
curl -s "$BASE/api/health" - Network accessline 207
curl -s "$BASE/api/accounts/acct_globex_north/metrics?start=2026-04&end=2026-06" - Network accessline 208
curl -s "$BASE/api/billing/snapshots?account_id=acct_globex_north&as_of=2026-06-30" - Network accessline 209
curl -s "$BASE/api/finance/ar-aging?as_of=2026-06-30" - Network accessline 210
curl -s "$BASE/exports/churn/candidates.csv"
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_004/self/self_attempt_01/SKILL.md
- Network accessline 7
A fresh solver gets ONE test prompt + the remote env. The prompt may name a local URL (`http://127.0.0.1:8074`) or an `env/setup.sh`. **IGNORE those.** Use ONLY the - Network accessline 8
URL (`http://127.0.0.1:8074`) or an `env/setup.sh`. **IGNORE those.** Use ONLY the - Network accessline 33
**Performance tip:** sequential per-account fan-out is slow; download all needed JSON to files with backgrounded/parallel `curl` (e.g. `curl ... -o f.json &` then - Network accessline 34
JSON to files with backgrounded/parallel `curl` (e.g. `curl ... -o f.json &` then
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_01/scripts/fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 7
Usage examples: ERP_BASE=http://127.0.0.1:8029 python scripts/fetch.py bills # dump all AP bills - Network accessline 8
ERP_BASE=http://127.0.0.1:8029 python scripts/fetch.py bills # dump all AP bills - Network accessline 21
[Credential-related source omitted] - Access to secretsline 21
[Credential-related source omitted] - Network accessline 24
def _get(path): with urllib.request.urlopen(BASE + path) as r: - Network accessline 25
with urllib.request.urlopen(BASE + path) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 3
description: >- Solve ERP finance close, AP, and vendor-compliance review tasks that read from the shared task_group_005 ERP finance JSON API (base http://127.0.0.1:8029). Use this WHENEVER a task asks - Network accessline 4
Solve ERP finance close, AP, and vendor-compliance review tasks that read from the shared task_group_005 ERP finance JSON API (base http://127.0.0.1:8029). Use this WHENEVER a task asks you to decide claim/AP/reimbursement close status, - Network accessline 5
task_group_005 ERP finance JSON API (base http://127.0.0.1:8029). Use this WHENEVER a task asks - Network accessline 26
1. **API is truth.** Base URL is always `http://127.0.0.1:8029`. If a prompt names another URL - Network accessline 44
Default `limit=100`, max `500`. Always pass `limit=500`; if `total > 500`, page with `offset`. Filter by any exact field name, e.g. `?status=paid`, `?claim_id=...`, `?business_id=...`, `?account=1250&period=2025-03`. Use `curl -s`. - Network accessline 46
`?account=1250&period=2025-03`. Use `curl -s`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 27
- **Base URL:** `http://127.0.0.1:8029`. **Ignore any other base URL** a prompt mentions (e.g. `:8005`, `:8029` variants in narrative text). Confirm with `GET /health`. - Fetch only via HTTP GET + `curl`. Never read local server files. - Network accessline 256
2. Read the payload (scope/batch IDs, as_of/period, thresholds) and the `answer_template.json` (keys, enums, ordering, precision). 3. Pull every needed record from the API at `http://127.0.0.1:8029` (paginate to `total`). - Network accessline 257
(keys, enums, ordering, precision). 3. Pull every needed record from the API at `http://127.0.0.1:8029` (paginate to `total`). 4. Apply the rules above; compute money to 2 decimals; dedupe and sort as specified. - Network accessline 258
3. Pull every needed record from the API at `http://127.0.0.1:8029` (paginate to `total`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_01/SKILL.md
- Network accessline 20
- **The live API is the system of record.** Any base URL written inside a prompt (e.g. `http://127.0.0.1:8005`) is wrong/stale — always use the runner-provided base URL. - Network accessline 21
(e.g. `http://127.0.0.1:8005`) is wrong/stale — always use the runner-provided base URL. Any local CSV/JSON "snapshot/export" is CONTEXT ONLY; never trust its statuses, amounts, or payment states. Re-fetch every record from the API and
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_005/self/self_attempt_03/SKILL.md
- Network accessline 17
- **The live API is the system of record.** Any base URL written in a prompt (e.g. `http://127.0.0.1:8005`) is a decoy — always use the runner-provided base URL. In this environment that base is `<remote-env-url>`. Query it with `curl`. - Network accessline 18
(e.g. `http://127.0.0.1:8005`) is a decoy — always use the runner-provided base URL. In this environment that base is `<remote-env-url>`. Query it with `curl`. - **Local payloads (CSV exports, snapshots, batch files) are CONTEXT, not trut
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 83
Base URL comes from the prompt (commonly `http://127.0.0.1:8056`, mirror at
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 62
### 2. Find the base URL and confirm the API is live The base URL is in the prompt (commonly `http://127.0.0.1:8006`; a mirror at - Network accessline 63
The base URL is in the prompt (commonly `http://127.0.0.1:8006`; a mirror at `http://127.0.0.1:8056` serves identical data — either works). Confirm with `GET /health` and skim `GET /manifest` for record counts and anchor ids. Use `curl` - Network accessline 64
`http://127.0.0.1:8056` serves identical data — either works). Confirm with `GET /health` and skim `GET /manifest` for record counts and anchor ids. Use `curl` or Python `urllib`/`requests`; all calls are read-only GETs. - Network accessline 99
A reusable helper is in `scripts/procureops.py` (a tiny GET client + rounding/sort helpers); import it or copy the snippet. It is optional — plain curl works too. - Network accessline 100
helpers); import it or copy the snippet. It is optional — plain curl works too.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02/references/api_field_map.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02/scripts/procureops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 12
Example: from procureops import Client, money, as_set api = Client("http://127.0.0.1:8006") - Network accessline 13
from procureops import Client, money, as_set api = Client("http://127.0.0.1:8006") po = api.get("/purchase_orders/PO-AX17-4481") - Network accessline 14
api = Client("http://127.0.0.1:8006") - Network accessline 29
class Client: def __init__(self, base_url: str = "http://127.0.0.1:8006", timeout: float = 15.0): - Network accessline 30
def __init__(self, base_url: str = "http://127.0.0.1:8006", timeout: float = 15.0): - Network accessline 39
if clean: url += "?" + urllib.parse.urlencode(clean) with urllib.request.urlopen(url, timeout=self.timeout) as resp: - Network accessline 40
url += "?" + urllib.parse.urlencode(clean) with urllib.request.urlopen(url, timeout=self.timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 41
with urllib.request.urlopen(url, timeout=self.timeout) as resp: - Network accessline 90
if __name__ == "__main__": # smoke test import sys base = sys.argv[1] if len(sys.argv) > 1 else "http://127.0.0.1:8006" - Network accessline 91
import sys base = sys.argv[1] if len(sys.argv) > 1 else "http://127.0.0.1:8006" api = Client(base) - Network accessline 92
base = sys.argv[1] if len(sys.argv) > 1 else "http://127.0.0.1:8006"
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_03/scripts/probe.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 8
Why this exists: every ProcureOps task starts the same way -- resolve a handful of ids against the live API, then walk PO -> receipts / invoices / payments and supplier -> risk events. Doing that by hand with curl is slow and error prone. - Network accessline 9
of ids against the live API, then walk PO -> receipts / invoices / payments and supplier -> risk events. Doing that by hand with curl is slow and error prone. This script gives you stable JSON you can pipe into your own arithmetic. - Network accessline 10
supplier -> risk events. Doing that by hand with curl is slow and error prone. - Network accessline 33
DEFAULT_BASE = "http://127.0.0.1:8056" - Network accessline 36
def _get(base, path): url = base.rstrip("/") + path with urllib.request.urlopen(url, timeout=20) as r: - Network accessline 37
url = base.rstrip("/") + path with urllib.request.urlopen(url, timeout=20) as r: return json.loads(r.read().decode("utf-8")) - Network accessline 38
with urllib.request.urlopen(url, timeout=20) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 17
`answer_template.json`. The prompt names the as-of date, the program, and the anchor PO/receipt/invoice IDs to focus on. 2. The prompt may print `http://127.0.0.1:8006`. Ignore that. Use the base URL - Network accessline 18
anchor PO/receipt/invoice IDs to focus on. 2. The prompt may print `http://127.0.0.1:8006`. Ignore that. Use the base URL the runner provides for the shared API. Treat it as read-only. - Network accessline 19
2. The prompt may print `http://127.0.0.1:8006`. Ignore that. Use the base URL
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/self/self_attempt_01/SKILL.md
- Network accessline 53
curl -s <remote-env-url>manifest - Network accessline 54
curl -s <remote-env-url><collection> # {"count":N,"results":[...]} - Network accessline 55
curl -s <remote-env-url><collection>/<id> # single object, 404 if absent - Network accessline 56
curl -s "<remote-env-url>receipts?po_id=PO-AX17-4481"
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_006/self/self_attempt_02/SKILL.md
- Network accessline 20
ERP yet (see Task-5 pattern). 2. **Base URL is supplied by the runner / environment_access.md.** The prompt's `http://127.0.0.1:8006` or `localhost:8006` is a placeholder — replace it with - Network accessline 21
2. **Base URL is supplied by the runner / environment_access.md.** The prompt's `http://127.0.0.1:8006` or `localhost:8006` is a placeholder — replace it with the real base URL from `environment_access.md` (e.g. - Network accessline 22
`http://127.0.0.1:8006` or `localhost:8006` is a placeholder — replace it with
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 31
Base URL: read `environment_access.md` in the task directory and use the **exact base URL it specifies** (currently `http://127.0.0.1:8015`). If task text mentions a different port (e.g. `8007`/`8015`), the access doc wins. Query with `curl
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_02/references/api_guide.md
- Network accessline 3
Read-only JSON over HTTP GET. **Base URL: use the exact one in `environment_access.md`** (currently `http://127.0.0.1:8015`). Ignore any other port mentioned in task prose. Filters are exact string match; omitting a filter returns all rows. - Network accessline 98
import urllib.request, json BASE = "http://127.0.0.1:8015" # confirm against environment_access.md - Network accessline 99
BASE = "http://127.0.0.1:8015" # confirm against environment_access.md def get(path): with urllib.request.urlopen(BASE + path) as r: - Network accessline 100
def get(path): with urllib.request.urlopen(BASE + path) as r: return json.load(r) - Network accessline 101
with urllib.request.urlopen(BASE + path) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 46
Use `http://127.0.0.1:8015`. **Ignore any other port** a memo or prompt mentions (e.g.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_03/references/api_reference.md
- Network accessline 3
Read-only JSON over plain HTTP GET. **Base URL: `http://127.0.0.1:8015`** (ignore any other - Network accessline 8
Query with `curl -s` or Python `urllib`. Example: - Network accessline 9
`curl -s "http://127.0.0.1:8015/orders?wave=TRAIN_TRANSFER_B"` - Network accessline 64
import urllib.request, json def get(path): return json.load(urllib.request.urlopen("http://127.0.0.1:8015" + path)) - Network accessline 66
return json.load(urllib.request.urlopen("http://127.0.0.1:8015" + path))
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_007/reflect-3/reflect-3_attempt_01/SKILL.md
- Network accessline 35
- Prefer `curl` for fetching; pull whole collections ONCE (`/products`, `/customers`, `/inventory`, `/purchase_orders`, `/suppliers`) and index them locally by key, rather than many per-id calls. (Some HTTP clients are very slow per-req - Network accessline 36
`/inventory`, `/purchase_orders`, `/suppliers`) and index them locally by key, rather than many per-id calls. (Some HTTP clients are very slow per-request; curl is fast.) - The remote can occasionally return an empty reply (curl exit 52). - Network accessline 37
than many per-id calls. (Some HTTP clients are very slow per-request; curl is fast.) - Network accessline 38
- The remote can occasionally return an empty reply (curl exit 52). Just retry the call.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 27
The template's `analysis_type` enum tells you which of the four task families you are in. 2. Resolve the API base URL. It is usually the env var `API_BASE`; otherwise default to `http://127.0.0.1:8066`. The API is GET-only. - Network accessline 28
2. Resolve the API base URL. It is usually the env var `API_BASE`; otherwise default to `http://127.0.0.1:8066`. The API is GET-only. 3. Pull every relevant record for the client (see §1) and resolve conflicting facts (§2). - Network accessline 29
`http://127.0.0.1:8066`. The API is GET-only.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_01/references/solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 10
Usage: python solver.py --selftest python solver.py --base http://127.0.0.1:8066 roth CLT-1001 --horizon 2046 - Network accessline 11
python solver.py --selftest python solver.py --base http://127.0.0.1:8066 roth CLT-1001 --horizon 2046 python solver.py --base http://127.0.0.1:8066 ilit CLT-1002 - Network accessline 12
python solver.py --base http://127.0.0.1:8066 roth CLT-1001 --horizon 2046 - Network accessline 13
python solver.py --base http://127.0.0.1:8066 ilit CLT-1002 - Network accessline 14
python solver.py --base http://127.0.0.1:8066 trust CLT-1003 - Network accessline 15
python solver.py --base http://127.0.0.1:8066 plan CLT-1004 - Network accessline 22
def api(base, path): with urllib.request.urlopen(base + path) as r: - Network accessline 23
with urllib.request.urlopen(base + path) as r: - Network accessline 213
ap.add_argument("family", nargs="?", choices=["roth", "ilit", "trust", "plan"]) ap.add_argument("client_id", nargs="?") ap.add_argument("--base", default=os.environ.get("API_BASE", "http://127.0.0.1:8066")) - Network accessline 214
ap.add_argument("client_id", nargs="?") ap.add_argument("--base", default=os.environ.get("API_BASE", "http://127.0.0.1:8066")) ap.add_argument("--horizon", type=int) - Network accessline 215
[Credential-related source omitted] - Access to secretsline 215
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 51
2. **Find the API base.** Use `API_BASE` env if set, else `http://127.0.0.1:8066`. Confirm with
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_02/reference.md
- Network accessline 21
Endpoints (all GET, base = `API_BASE` env or `http://127.0.0.1:8066`):
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_02/scripts/advisory_lib.py
- Executable scriptline 1
Executable file - Network accessline 15
All HTTP is GET-only. Base URL is given by the harness (env API_BASE, else http://127.0.0.1:8066). - Network accessline 16
http://127.0.0.1:8066). - Network accessline 23
[Credential-related source omitted] - Access to secretsline 23
[Credential-related source omitted] - Network accessline 28
raw = subprocess.check_output(["curl", "-s", BASE + path])
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 30
`estate_liquidity_action_plan`. 3. Pull every fact from the **API**, never from your own knowledge. Base URL is whatever the harness exposes (env `API_BASE`, usually `http://127.0.0.1:8066`). - Network accessline 31
3. Pull every fact from the **API**, never from your own knowledge. Base URL is whatever the harness exposes (env `API_BASE`, usually `http://127.0.0.1:8066`). Tax constants and RMD factors come from `/api/policies/tax` and - Network accessline 32
whatever the harness exposes (env `API_BASE`, usually `http://127.0.0.1:8066`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/demonstration_skill/demonstration_skill_attempt_03/scripts/advisory_lib.py
- Executable scriptline 1
Executable file - Network accessline 25
[Credential-related source omitted] - Access to secretsline 25
[Credential-related source omitted] - Network accessline 32
@lru_cache(maxsize=None) def fetch(path): - Network accessline 33
def fetch(path): - Network accessline 34
with urllib.request.urlopen(BASE + path) as r: - Network accessline 43
def tax_policy(): return fetch("/api/policies/tax") - Network accessline 44
return fetch("/api/policies/tax") - Network accessline 47
def rmd_factors(): return {int(k): v for k, v in fetch("/api/rmd-factors").items()} - Network accessline 48
return {int(k): v for k, v in fetch("/api/rmd-factors").items()} - Network accessline 51
def client_header(cid): return fetch(f"/api/clients/{cid}") - Network accessline 52
return fetch(f"/api/clients/{cid}") - Network accessline 57
out = {} for d in fetch(f"/api/source-documents?client_id={cid}"): - Network accessline 58
for d in fetch(f"/api/source-documents?client_id={cid}"): - Network accessline 111
pol = tax_policy() rmd = rmd_factors() acct = first_or_none(fetch(f"/api/retirement-accounts?client_id={cid}")) - Network accessline 113
acct = first_or_none(fetch(f"/api/retirement-accounts?client_id={cid}")) - Network accessline 211
pol = tax_policy() ec = estate_context(cid) policy = first_or_none(fetch(f"/api/life-insurance?client_id={cid}")) - Network accessline 212
ec = estate_context(cid) policy = first_or_none(fetch(f"/api/life-insurance?client_id={cid}")) planning_year = ec["planning_year"] - Network accessline 213
policy = first_or_none(fetch(f"/api/life-insurance?client_id={cid}")) - Network accessline 311
def trust_numbers(cid): tc = first_or_none(fetch(f"/api/trust-candidates?client_id={cid}")) - Network accessline 312
tc = first_or_none(fetch(f"/api/trust-candidates?client_id={cid}"))
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_01/SKILL.md
- Network accessline 34
## 1. API access Base URL is given by the harness (env `API_BASE`, or in `environment_access.md`). At time of writing: `<remote-env-url>`. Use `curl -s` via Bash. - Network accessline 35
Base URL is given by the harness (env `API_BASE`, or in `environment_access.md`). At time of writing: `<remote-env-url>`. Use `curl -s` via Bash. If you run helper scripts use `python` (NOT `python3` — broken on the host). - Network accessline 36
At time of writing: `<remote-env-url>`. Use `curl -s` via Bash. - Network accessline 315
## 8. Procedure (SOP) 1. `curl /api/health` to confirm liveness; read base URL from environment. - Network accessline 316
1. `curl /api/health` to confirm liveness; read base URL from environment.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_02/SKILL.md
- Network accessline 42
Use `curl -s` via Bash. If you run helper scripts use `python` (not `python3`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_008/self/self_attempt_03/SKILL.md
- Network accessline 47
## 1. API usage (read-only, curl via Bash) - Network accessline 49
Base URL comes from the environment (the harness exposes it; the file `environment_access.md` gives it explicitly, e.g. `<remote-env-url>`). All endpoints are GET and return JSON. Use `curl -s --max-time 30`. - Network accessline 51
All endpoints are GET and return JSON. Use `curl -s --max-time 30`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 32
`required_top_level_keys`, the `field_types`, the rounding rules in `description`, and any ordering rules. Build exactly those keys — no more. 4. Query the API (curl or python). Compute. Emit JSON. - Network accessline 34
4. Query the API (curl or python). Compute. Emit JSON. - Network accessline 41
import sys; sys.path.insert(0, "scripts") from finance_ops_helpers import * BASE = "http://127.0.0.1:8028" # always read the real value from the payload - Network accessline 43
BASE = "http://127.0.0.1:8028" # always read the real value from the payload
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_01/scripts/finance_ops_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 10
Usage: from finance_ops_helpers import * BASE = "http://127.0.0.1:8028" # read from payloads/environment_access.json - Network accessline 11
from finance_ops_helpers import * BASE = "http://127.0.0.1:8028" # read from payloads/environment_access.json """ - Network accessline 12
BASE = "http://127.0.0.1:8028" # read from payloads/environment_access.json - Network accessline 23
with urllib.request.urlopen(base_url.rstrip("/") + path) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 36
Read `payloads/answer_template.json` for the exact output keys, field types, and ordering rules. 2. Query the API with curl or python (GET only). Look at the real shapes first; - Network accessline 37
and ordering rules. 2. Query the API with curl or python (GET only). Look at the real shapes first; do not assume field names. - Network accessline 38
2. Query the API with curl or python (GET only). Look at the real shapes first;
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_02/scripts/finance_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 25
BASE = "http://127.0.0.1:8028" # read base_url from payloads/environment_access.json - Network accessline 28
def get(path): return json.load(urllib.request.urlopen(BASE + path)) - Network accessline 29
return json.load(urllib.request.urlopen(BASE + path))
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/demonstration_skill/demonstration_skill_attempt_03/scripts/finance_ops.py
- Executable scriptline 1
Executable file - Network accessline 17
BASE_URL = "http://127.0.0.1:8028" # confirm against payloads/environment_access.json - Network accessline 20
def get(path, base=BASE_URL): with urllib.request.urlopen(base + path, timeout=20) as r: - Network accessline 21
with urllib.request.urlopen(base + path, timeout=20) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_01/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_02/SKILL.md
- Network accessline 21
to emit (no more, no less). `field_types` defines every nested key and its shape. Match key names, nesting, and ordering EXACTLY. The `description` field states rounding and list-ordering rules. - `payloads/environment_access.json` — IGNO - Network accessline 22
nesting, and ordering EXACTLY. The `description` field states rounding and list-ordering rules. - `payloads/environment_access.json` — IGNORE its `base_url` (e.g. `http://127.0.0.1:...`). It is wrong. Always use the remote base URL below. - Network accessline 23
- `payloads/environment_access.json` — IGNORE its `base_url` (e.g. `http://127.0.0.1:...`). It is wrong. - Network accessline 28
- Base URL: `<remote-env-url>` - Fetch with HTTP GET, e.g. `curl -s "<remote-env-url>api/finance/records?branch_id=BR-004"`. - Network accessline 29
- Fetch with HTTP GET, e.g. `curl -s "<remote-env-url>api/finance/records?branch_id=BR-004"`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/reflect-3/reflect-3_attempt_03/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_01/SKILL.md
- Network accessline 48
Base URL: `<remote-env-url>` (ALWAYS use this; ignore any `base_url` like `http://127.0.0.1:8047` inside `payloads/environment_access.json`). All endpoints are HTTP GET. - Network accessline 49
`http://127.0.0.1:8047` inside `payloads/environment_access.json`). All endpoints are HTTP GET. - Network accessline 65
Fetch once into files, then compute in code (Python). On this Windows/Git-Bash host write temp files under the session scratchpad, not `/tmp`. Example: `curl -s "<remote-env-url>api/finance/records?branch_id=BR-004" -o records.json` - Network accessline 67
`curl -s "<remote-env-url>api/finance/records?branch_id=BR-004" -o records.json`
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_02/SKILL.md
- Network accessline 22
mean: trust ONLY the live API data and ignore any narrative figures. Reconciliations tie to zero (see region task). 5. Fetch with `curl -s`. All endpoints are GET. Compute in a script, then format. - Network accessline 24
5. Fetch with `curl -s`. All endpoints are GET. Compute in a script, then format.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_009/self/self_attempt_03/SKILL.md
- Network accessline 13
2. Read `request_memo.json` for the **entity id(s)** (`target_branch_id`, `target_region_id`, `ensemble_id`, `scenario_id`, `production_id`), period fields, and the `*_focus` list (focus lists tell you which computations matter; they do not - Network accessline 14
3. Read `answer_template.json`. The `required_top_level_keys` list is the **exact** set of top-level keys your object must have (no more, no less). `field_types` gives the nested shape, the enum domains, and the rounding hints. Build your o - Network accessline 15
4. Ignore any `base_url` inside payloads (e.g. `http://127.0.0.1:...`). **Always** use the remote base URL below.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 28
or preferences. Note the `portfolio_id`, the requested universe/quarter/window, and the exact output contract in `answer_template.json`. 2. **Pull live data from the API** (base `http://127.0.0.1:8036`, all `GET`): - Network accessline 29
and the exact output contract in `answer_template.json`. 2. **Pull live data from the API** (base `http://127.0.0.1:8036`, all `GET`): start with `/api/policies` (thresholds), `/api/catalog` (valid ids), and the - Network accessline 30
2. **Pull live data from the API** (base `http://127.0.0.1:8036`, all `GET`):
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 29
preferences and thresholds — but treat its marks/holdings/ratings/watchlist/dates as possibly stale. 3. **Pull current truth from the API.** Base URL `http://127.0.0.1:8036`. All endpoints are - Network accessline 30
possibly stale. 3. **Pull current truth from the API.** Base URL `http://127.0.0.1:8036`. All endpoints are `GET` returning JSON. Endpoints: - Network accessline 31
3. **Pull current truth from the API.** Base URL `http://127.0.0.1:8036`. All endpoints are
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 49
## 1. Environment API quick reference (base `http://127.0.0.1:8036`, all GET/JSON)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 3
description: >- Produce committee-ready JSON answers for the shared "credit office" lending-committee task family (bank branches and credit-union segments served by the HTTP API at http://127.0.0.1:8003). Use this - Network accessline 4
Produce committee-ready JSON answers for the shared "credit office" lending-committee task family (bank branches and credit-union segments served by the HTTP API at http://127.0.0.1:8003). Use this skill whenever a prompt asks you to pr - Network accessline 5
(bank branches and credit-union segments served by the HTTP API at http://127.0.0.1:8003). Use this - Network accessline 44
Base URL: `http://127.0.0.1:8003`. All JSON; money/ratios are plain numbers.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
- Network accessline 3
description: > Produce committee-ready JSON answers for the shared Credit Office / lending-committee task family backed by the HTTP API at http://127.0.0.1:8003. Use this skill WHENEVER a - Network accessline 4
Produce committee-ready JSON answers for the shared Credit Office / lending-committee task family backed by the HTTP API at http://127.0.0.1:8003. Use this skill WHENEVER a prompt asks you to prepare a credit-risk, lending-committee, or - Network accessline 5
task family backed by the HTTP API at http://127.0.0.1:8003. Use this skill WHENEVER a - Network accessline 29
exactly the keys the template lists, in the template's structure. Extra keys are risk; missing required keys fail. 2. **The API is the only data source.** Use `curl`/`urllib` against `http://127.0.0.1:8003`. - Network accessline 30
missing required keys fail. 2. **The API is the only data source.** Use `curl`/`urllib` against `http://127.0.0.1:8003`. Ignore any mention of `env/setup.sh`, local DBs, or files — the service is already up. - Network accessline 31
2. **The API is the only data source.** Use `curl`/`urllib` against `http://127.0.0.1:8003`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_02/references/policy.md
- Network accessline 9
Base: `http://127.0.0.1:8003`. All JSON. `branch_id`/`segment_id` matched case-insensitively.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_02/scripts/credit_office.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 8
Usage: from credit_office import API, Policy api = API() # http://127.0.0.1:8003 - Network accessline 9
from credit_office import API, Policy api = API() # http://127.0.0.1:8003 pol = Policy(api.policies()) - Network accessline 10
api = API() # http://127.0.0.1:8003 - Network accessline 24
BASE = "http://127.0.0.1:8003" - Network accessline 48
if params: url += "?" + urllib.parse.urlencode({k: v for k, v in params.items() if v is not None}) with urllib.request.urlopen(url) as r: - Network accessline 49
url += "?" + urllib.parse.urlencode({k: v for k, v in params.items() if v is not None}) with urllib.request.urlopen(url) as r: return json.load(r) - Network accessline 50
with urllib.request.urlopen(url) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/demonstration_skill/demonstration_skill_attempt_03/SKILL.md
- Network accessline 11
variance (FDIC/NCUA), sector/CRE concentration limits, lending capacity allocation, or asks for a JSON answer matching an answer_template.json against the credit office HTTP API at http://127.0.0.1:8003. Apply it even when the prompt on - Network accessline 12
or asks for a JSON answer matching an answer_template.json against the credit office HTTP API at http://127.0.0.1:8003. Apply it even when the prompt only hints at these (e.g. "regrade the loans rated 3 or worse", "compare the two CRE r - Network accessline 13
HTTP API at http://127.0.0.1:8003. Apply it even when the prompt only hints at these (e.g. "regrade the loans rated 3 or worse", "compare the two CRE requests", "posture recommendation for the segment") and does not name the skill expli - Network accessline 41
If the prompt mentions `env/setup.sh` or an "API base URL printed by setup", ignore that plumbing — the API is already live at `http://127.0.0.1:8003`. Do not look for local - Network accessline 42
plumbing — the API is already live at `http://127.0.0.1:8003`. Do not look for local - Network accessline 62
Base URL `http://127.0.0.1:8003`. All responses are JSON; money/ratios are plain numbers. `branch_id` and `segment_id` are matched case-insensitively. Fetch with `curl` or Python - Network accessline 221
**interpretation:** `capacity_status` (`capacity_available` when `quarterly_capacity` remains), `external_risk_status` (`weaker_than_national_and_peers` when NC is worse on the - Network accessline 222
remains), `external_risk_status` (`weaker_than_national_and_peers` when NC is worse on the
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/self/self_attempt_01/SKILL.md
- Network accessline 30
Base URL: `<remote-env-url>`. Use `curl -s` (Bash tool). The environment - Network accessline 150
**integers exactly as reported**: `delinquency_bps`, `loan_to_share_pct`, `roaa_bps`, `positive_net_income_pct`. Peer states come from the segment's `peer_states` list. - direction = NC value vs comparison: `higher` if NC>comp, `lower` if N - Network accessline 151
`positive_net_income_pct`. Peer states come from the segment's `peer_states` list. - direction = NC value vs comparison: `higher` if NC>comp, `lower` if NC<comp, `equal`. - Compare NC to `US` row (`nc_vs_us`) and to the **median of peer-sta - Network accessline 152
- direction = NC value vs comparison: `higher` if NC>comp, `lower` if NC<comp, `equal`. - Network accessline 153
- Compare NC to `US` row (`nc_vs_us`) and to the **median of peer-state values per metric** - Network accessline 155
- Polarity: for delinquency, lower is better (higher = weaker); for roaa_bps, loan_to_share_pct (within prudent range), positive_net_income_pct, higher is generally stronger. Net "external_risk_status" reflects whether NC is broadly wea - Network accessline 157
stronger. Net "external_risk_status" reflects whether NC is broadly weaker/stronger/mixed.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_011/self/self_attempt_03/SKILL.md
- Network accessline 233
positive_net_income_pct). - `peer_comparison.peer_states` = segment `peer_states`, **sorted ascending**. `nc_vs_us` = direction of NC value vs the `US` row per metric - Network accessline 234
- `peer_comparison.peer_states` = segment `peer_states`, **sorted ascending**. `nc_vs_us` = direction of NC value vs the `US` row per metric (higher/lower/equal). `nc_vs_peer_median` = direction vs the **median of the - Network accessline 235
`nc_vs_us` = direction of NC value vs the `US` row per metric - Network accessline 238
worse, lower roaa / lower positive_net_income = worse; loan_to_share is utilization (report direction, not "good/bad"). - `external_risk_status`: if NC is worse than both US and peers on the risk/ - Network accessline 239
utilization (report direction, not "good/bad"). - `external_risk_status`: if NC is worse than both US and peers on the risk/ earnings metrics → `weaker_than_national_and_peers`; mixed → `mixed_...`; - Network accessline 240
- `external_risk_status`: if NC is worse than both US and peers on the risk/
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 34
2. **Reach the API.** Read the task's `environment_access.md` for the base URL, then `GET /health` to confirm it is up. The web-UI login is not needed — call the JSON API directly with `curl -s`. - Network accessline 35
`GET /health` to confirm it is up. The web-UI login is not needed — call the JSON API directly with `curl -s`. 3. **Identify the task shape** from the prompt and follow the matching playbook in - Network accessline 36
API directly with `curl -s`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_01/references/api.md
- Network accessline 3
The environment is a local HR-lifecycle portal. Reach it ONLY over HTTP at the base URL given in the task's `environment_access.md` (it specifies the host/port, e.g. `http://127.0.0.1:<port>/`). The JSON API needs no auth; the login in the - Network accessline 4
base URL given in the task's `environment_access.md` (it specifies the host/port, e.g. `http://127.0.0.1:<port>/`). The JSON API needs no auth; the login in the prompt is for the web UI only. Always confirm reachability first with `GET /hea - Network accessline 5
e.g. `http://127.0.0.1:<port>/`). The JSON API needs no auth; the login in the - Network accessline 8
Use `curl -s "<base>/api/..."`. The `q` parameter is a case-insensitive substring
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_012/self/self_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_01/scripts/portal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 57
headers = {"Content-Type": "application/json"} if data is not None else {} body = json.dumps(data).encode() if data is not None else None req = urllib.request.Request(url, data=body, headers=headers, method=method) - Network accessline 58
body = json.dumps(data).encode() if data is not None else None req = urllib.request.Request(url, data=body, headers=headers, method=method) try: - Network accessline 59
req = urllib.request.Request(url, data=body, headers=headers, method=method) - Network accessline 60
try: with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode()) - Network accessline 61
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_01/scripts/template_constraints.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_02/references/portal_api.md
- Network accessline 11
BASE=$GDPEVO_ENV_BASE_URL # e.g. http://task-env:9013/ curl -s "$BASE" # HTML landing page (confirms reachability) - Network accessline 12
curl -s "$BASE" # HTML landing page (confirms reachability) - Network accessline 13
curl -s "$BASE/health" # liveness - Network accessline 39
q() { curl -s -X POST "$BASE/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_03/references/environment.md
- Network accessline 3
The task base URL is given in `environment_access.md` as `GDPEVO_ENV_BASE_URL` (e.g. `http://task-env:9013/`). No credentials. Substitute it for every - Network accessline 4
(e.g. `http://task-env:9013/`). No credentials. Substitute it for every - Network accessline 39
Curl pattern used throughout: - Network accessline 41
curl -s -X POST <BASE>/query -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_01/references/portal_api.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_01/scripts/portal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 32
[Credential-related source omitted] - Network accessline 52
body = json.dumps(data).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, headers=headers, - Network accessline 53
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, headers=headers, method="POST" if data is not None else "GET") - Network accessline 54
req = urllib.request.Request(url, data=body, headers=headers, - Network accessline 55
method="POST" if data is not None else "GET") with urllib.request.urlopen(req, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 56
with urllib.request.urlopen(req, timeout=timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_02/SKILL.md
- Network accessline 89
curl -s -X POST "$BASE/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_03/SKILL.md
- Network accessline 46
2. **Resolve the base URL.** The prompt uses the placeholder `<TASK_ENV_BASE_URL>`. The real base URL is in `environment_access.md` (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9013/`). Use **only** the - Network accessline 47
`<TASK_ENV_BASE_URL>`. The real base URL is in `environment_access.md` (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9013/`). Use **only** the endpoints listed there. No credentials are needed. - Network accessline 48
(`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9013/`). Use **only** the
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_013/self/self_attempt_03/references/portal.md
- Network accessline 33
curl -s -X POST "$BASE/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_01/scripts/date_add.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_01/scripts/nsql.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 7
Usage: python3 nsql.py "SELECT * FROM cases WHERE case_id='CASE-X'" python3 nsql.py --base http://task-env:9014 --token pa-review-token-014 "PRAGMA table_info(cases)" - Network accessline 9
python3 nsql.py --base http://task-env:9014 --token pa-review-token-014 "PRAGMA table_info(cases)" - Access to secretsline 12
[Credential-related source omitted] - Access to secretsline 29
[Credential-related source omitted] - Access to secretsline 31
[Credential-related source omitted] - Access to secretsline 32
[Credential-related source omitted] - Access to secretsline 33
[Credential-related source omitted] - Network accessline 43
url = args.base.rstrip("/") + "/sql/query" body = json.dumps({"sql": args.sql}).encode("utf-8") req = urllib.request.Request(url, data=body, method="POST") - Network accessline 44
body = json.dumps({"sql": args.sql}).encode("utf-8") req = urllib.request.Request(url, data=body, method="POST") req.add_header("Content-Type", "application/json") - Network accessline 45
req = urllib.request.Request(url, data=body, method="POST") - Network accessline 50
try: with urllib.request.urlopen(req, timeout=args.timeout) as resp: - Network accessline 51
with urllib.request.urlopen(req, timeout=args.timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 55
- Base URL: the value of `<TASK_ENV_BASE_URL>` (e.g. `http://task-env:9014/`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_02/reference/environment.md
- Network accessline 10
- Base URL: value of `<TASK_ENV_BASE_URL>` (observed: `http://task-env:9014/`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/fewshot/fewshot_attempt_03/scripts/nsql.sh
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/self/self_attempt_01/reference/schema.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_014/self/self_attempt_02/references/environment_schema.md
- Network accessline 37
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 42
3. **Connect.** Read `environment_access.md` for `GDPEVO_ENV_BASE_URL` and the allow-listed endpoints. It is HTTP, no auth. Only GET the listed paths. Example: `curl -sS "$BASE/api/patients/{id}/conditions" | jq .` - Network accessline 43
allow-listed endpoints. It is HTTP, no auth. Only GET the listed paths. Example: `curl -sS "$BASE/api/patients/{id}/conditions" | jq .` 4. **Gather evidence.** For each case object, GET its detail plus every related - Network accessline 44
`curl -sS "$BASE/api/patients/{id}/conditions" | jq .`
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_02/references/data-model.md
- Network accessline 9
`{"referrals":[…]}`, `{"service_requests":[…]}`, `{"audit_logs":[…]}`); per-patient clinical lists return a bare JSON array. When in doubt, inspect the top-level keys first (`curl … | jq 'keys'`). - Network accessline 11
first (`curl … | jq 'keys'`). - Network accessline 80
- `curl -sS "$BASE/api/…" | jq .` to inspect; `jq -c '.[]?'` to stream rows.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_03/scripts/ehr_get.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 26
[Credential-related source omitted] - Network accessline 41
def get(base_url, path, timeout): url = base_url + ("" if path.startswith("/") else "/") + path req = urllib.request.Request(url, method="GET") - Network accessline 42
url = base_url + ("" if path.startswith("/") else "/") + path req = urllib.request.Request(url, method="GET") try: - Network accessline 43
req = urllib.request.Request(url, method="GET") - Network accessline 44
try: with urllib.request.urlopen(req, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 45
with urllib.request.urlopen(req, timeout=timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/self/self_attempt_02/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_015/self/self_attempt_03/references/api_reference.md
- Network accessline 49
- Prefer `curl -s` + `jq`; guard for both the wrapped and bare shapes.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_01/references/environment_api.md
- Network accessline 4
field names, not answers — always read live data and the live protocol body for the case in front of you. The base URL comes from `environment_access.md` (`GDPEVO_ENV_BASE_URL`), e.g. `http://task-env:9016/`. - Network accessline 6
(`GDPEVO_ENV_BASE_URL`), e.g. `http://task-env:9016/`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_01/scripts/gather_case.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 14
Example: python3 gather_case.py http://task-env:9016 CASE-RESP-102 - Network accessline 15
python3 gather_case.py http://task-env:9016 CASE-RESP-102 - Network accessline 35
url = base.rstrip("/") + path try: with urllib.request.urlopen(url, timeout=15) as resp: - Network accessline 36
try: with urllib.request.urlopen(url, timeout=15) as resp: return json.load(resp) - Network accessline 37
with urllib.request.urlopen(url, timeout=15) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_03/scripts/fetch_clinic_case.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 14
The base URL defaults to the GDPEVO_ENV_BASE_URL environment variable (as given in the run's environment-access file). Example: GDPEVO_ENV_BASE_URL=http://task-env:9016/ python3 fetch_clinic_case.py CASE-K-303 - Network accessline 15
in the run's environment-access file). Example: GDPEVO_ENV_BASE_URL=http://task-env:9016/ python3 fetch_clinic_case.py CASE-K-303 """ - Network accessline 16
GDPEVO_ENV_BASE_URL=http://task-env:9016/ python3 fetch_clinic_case.py CASE-K-303 - Network accessline 35
def get(base, path, timeout=15): url = base.rstrip("/") + path req = urllib.request.Request(url, method="GET") - Network accessline 36
url = base.rstrip("/") + path req = urllib.request.Request(url, method="GET") try: - Network accessline 37
req = urllib.request.Request(url, method="GET") - Network accessline 38
try: with urllib.request.urlopen(req, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(req, timeout=timeout) as resp: - Network accessline 52
"--base", default=os.environ.get("GDPEVO_ENV_BASE_URL", ""), help="Base URL (default: $GDPEVO_ENV_BASE_URL)", - Network accessline 53
[Credential-related source omitted] - Access to secretsline 53
[Credential-related source omitted] - Network accessline 54
help="Base URL (default: $GDPEVO_ENV_BASE_URL)",
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/fewshot/fewshot_attempt_03_invalid_20260725T215503Z/SKILL.md
- Network accessline 159
export GDPEVO_ENV_BASE_URL=http://task-env:9016/
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_016/self/self_attempt_01/references/data-model.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_01/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_01/scripts/pull_matter.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 5
[Credential-related source omitted] - Access to secretsline 5
[Credential-related source omitted] - Network accessline 6
# Example: # pull_matter.sh http://task-env:9017 review-key-017 MTR-EXAMPLE-GJ ./hub_dump # - Network accessline 7
# pull_matter.sh http://task-env:9017 review-key-017 MTR-EXAMPLE-GJ ./hub_dump - Network accessline 35
[Credential-related source omitted] - Network accessline 51
[Credential-related source omitted] - Network accessline 52
[Credential-related source omitted] - Network accessline 53
[Credential-related source omitted] - Network accessline 56
[Credential-related source omitted] - Network accessline 57
[Credential-related source omitted] - Network accessline 58
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 59
- Read the base URL + API key from `environment_access.md` / the matter context payload for the current run; do not assume. (In the environment observed while building this skill the base URL was `http://task-env:9017/` and the header w - Network accessline 60
[Credential-related source omitted] - Network accessline 61
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_03/scripts/pull_matter.sh
- Executable scriptline 1
#!/usr/bin/env bash - Access to secretsline 7
[Credential-related source omitted] - Network accessline 8
# present, or fall back to the values passed as flags. # # ./pull_matter.sh -m MTR-EXAMPLE-GJ [-b http://host:port] [-k api-key] [-e path/to/environment_access.md] - Network accessline 9
# # ./pull_matter.sh -m MTR-EXAMPLE-GJ [-b http://host:port] [-k api-key] [-e path/to/environment_access.md] # - Network accessline 10
# ./pull_matter.sh -m MTR-EXAMPLE-GJ [-b http://host:port] [-k api-key] [-e path/to/environment_access.md] - Access to secretsline 17
[Credential-related source omitted] - Network accessline 49
[Credential-related source omitted] - Network accessline 50
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/reflect-3/reflect-3_attempt_02/reference/pipeline.md
- Access to secretsline 25
[Credential-related source omitted] - Access to secretsline 26
[Credential-related source omitted] - Access to secretsline 27
[Credential-related source omitted] - Network accessline 33
def query(sql): req = urllib.request.Request(BASE + QPATH, - Network accessline 34
req = urllib.request.Request(BASE + QPATH, - Network accessline 35
data=json.dumps({"sql": sql}).encode(), headers=HDR, method="POST") with urllib.request.urlopen(req, timeout=20) as r: return json.load(r)["rows"] - Network accessline 36
with urllib.request.urlopen(req, timeout=20) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_017/self/self_attempt_03/scripts/hub.sh
- Executable scriptline 1
#!/usr/bin/env sh - Access to secretsline 8
[Credential-related source omitted] - Access to secretsline 22
[Credential-related source omitted] - Network accessline 26
if [ -z "$b" ]; then case "${1:-}" in http://*|https://*) b="$1" ;; - Network accessline 27
case "${1:-}" in http://*|https://*) b="$1" ;; *) echo "hub.sh: no base URL (set GDPEVO_ENV_BASE_URL or HUB_BASE_URL)" >&2; exit 2 ;; - Network accessline 28
http://*|https://*) b="$1" ;; - Network accessline 44
[Credential-related source omitted] - Network accessline 45
[Credential-related source omitted] - Network accessline 46
[Credential-related source omitted] - Network accessline 51
[Credential-related source omitted] - Network accessline 52
[Credential-related source omitted] - Network accessline 53
[Credential-related source omitted] - Network accessline 57
[Credential-related source omitted] - Network accessline 58
[Credential-related source omitted] - Network accessline 59
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_01/references/portal_api.md
- Network accessline 3
The portal is the **authoritative case-management system (CMS)** and current rule tables. Read the running environment's `environment_access.md` for the base URL (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9018/`) and the list of - Network accessline 4
rule tables. Read the running environment's `environment_access.md` for the base URL (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9018/`) and the list of allowed endpoints. Credentials are `none`. The prompt substitutes the real base - Network accessline 5
URL (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9018/`) and the list of - Network accessline 23
does **not** index citation/petition numbers — use the exact filter for those. The param is `q` (empty/`query=` returns nothing). - Curl pattern: - Network accessline 24
The param is `q` (empty/`query=` returns nothing). - Curl pattern: `curl -s "$BASE/api/cases?case_number=<CASE>" | python3 -m json.tool` - Network accessline 25
- Curl pattern: - Dangerous shell operationline 26
`curl -s "$BASE/api/cases?case_number=<CASE>" | python3 -m json.tool` - Network accessline 26
`curl -s "$BASE/api/cases?case_number=<CASE>" | python3 -m json.tool`
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_01/scripts/installments.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 210
- `reference/reconciliation_rules.md` — the full conflict/trap catalog with the reasoning behind each resolution. - `scripts/portal.sh` — curl wrapper for the portal. - Network accessline 211
reasoning behind each resolution. - `scripts/portal.sh` — curl wrapper for the portal. - `scripts/installments.py` — installment schedule + date calculator. - Network accessline 212
- `scripts/portal.sh` — curl wrapper for the portal.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_02/scripts/installments.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_02/scripts/portal.sh
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/fewshot/fewshot_attempt_03/scripts/finance_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/reflect-3/reflect-3_attempt_01/scripts/installments.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_018/self/self_attempt_01/reference/portal_api.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_01/references/environment.md
- Network accessline 8
- `GDPEVO_ENV_BASE_URL` — base URL, e.g. `http://task-env:9019/`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_01/scripts/env_query.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 33
get) path="${arg#/}" curl -sS -m 30 "${BASE}/${path}" - Network accessline 34
path="${arg#/}" curl -sS -m 30 "${BASE}/${path}" ;; - Network accessline 35
curl -sS -m 30 "${BASE}/${path}" - Network accessline 39
body="$(jq -nc --arg q "$arg" '{query:$q}')" - Network accessline 40
curl -sS -m 30 -X POST "${BASE}/api/sql" \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_02/scripts/licensing_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 71
e = env() url = e["base_url"] + (path if path.startswith("/") else "/" + path) with urllib.request.urlopen(url, timeout=30) as r: - Network accessline 72
url = e["base_url"] + (path if path.startswith("/") else "/" + path) with urllib.request.urlopen(url, timeout=30) as r: data = json.loads(r.read().decode()) - Network accessline 73
with urllib.request.urlopen(url, timeout=30) as r: - Network accessline 87
e = env() body = json.dumps({"query": query}).encode() req = urllib.request.Request( - Network accessline 88
body = json.dumps({"query": query}).encode() req = urllib.request.Request( e["base_url"] + "/api/sql", - Network accessline 89
req = urllib.request.Request( - Network accessline 93
method="POST", ) with urllib.request.urlopen(req, timeout=30) as r: - Network accessline 94
) with urllib.request.urlopen(req, timeout=30) as r: payload = json.loads(r.read().decode()) - Network accessline 95
with urllib.request.urlopen(req, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 31
- `GDPEVO_ENV_BASE_URL` — base URL for all requests (e.g. `http://task-env:9019/`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03/reference/contractor_review.py
- Executable scriptline 1
Executable file
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03/reference/env_client.py
- Executable scriptline 1
Executable file - Network accessline 31
if not path.startswith("/"): path = "/" + path with urllib.request.urlopen(self.base + path, timeout=30) as r: - Network accessline 32
path = "/" + path with urllib.request.urlopen(self.base + path, timeout=30) as r: return json.loads(r.read()) - Network accessline 33
with urllib.request.urlopen(self.base + path, timeout=30) as r: - Network accessline 38
body = json.dumps({"query": query}).encode() req = urllib.request.Request( - Network accessline 39
req = urllib.request.Request( - Network accessline 41
[Credential-related source omitted] - Network accessline 42
[Credential-related source omitted] - Network accessline 43
with urllib.request.urlopen(req, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03/reference/liquor_package.py
- Executable scriptline 1
Executable file
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/reflect-3/reflect-3_attempt_03/reference/renewal_queue.py
- Executable scriptline 1
Executable file
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/self/self_attempt_01/scripts/query_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 66
base, _ = load_env(env_file) url = base + "/api/" + path.lstrip("/").removeprefix("api/") with urllib.request.urlopen(url, timeout=30) as resp: - Network accessline 67
url = base + "/api/" + path.lstrip("/").removeprefix("api/") with urllib.request.urlopen(url, timeout=30) as resp: return json.load(resp) - Network accessline 68
with urllib.request.urlopen(url, timeout=30) as resp: - Network accessline 77
) body = json.dumps({"query": query}).encode("utf-8") req = urllib.request.Request( - Network accessline 78
body = json.dumps({"query": query}).encode("utf-8") req = urllib.request.Request( base + "/api/sql", - Network accessline 79
req = urllib.request.Request( - Network accessline 83
method="POST", ) with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 84
) with urllib.request.urlopen(req, timeout=30) as resp: return json.load(resp) - Network accessline 85
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_019/self/self_attempt_03/scripts/licensing_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 47
[Credential-related source omitted] - Access to secretsline 48
[Credential-related source omitted] - Network accessline 78
def http_get(base, endpoint): url = base + ("" if endpoint.startswith("/") else "/") + endpoint req = urllib.request.Request(url, method="GET") - Network accessline 79
url = base + ("" if endpoint.startswith("/") else "/") + endpoint req = urllib.request.Request(url, method="GET") with urllib.request.urlopen(req, timeout=30) as r: - Network accessline 80
req = urllib.request.Request(url, method="GET") - Network accessline 81
with urllib.request.urlopen(req, timeout=30) as r: - Network accessline 90
url = base + "/api/sql" body = json.dumps({"query": query}).encode("utf-8") req = urllib.request.Request( - Network accessline 91
body = json.dumps({"query": query}).encode("utf-8") req = urllib.request.Request( url, - Network accessline 92
req = urllib.request.Request( - Network accessline 96
[Credential-related source omitted] - Network accessline 97
) with urllib.request.urlopen(req, timeout=30) as r: return json.loads(r.read().decode("utf-8")) - Network accessline 98
with urllib.request.urlopen(req, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_01/reference/workbench_data_model.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_01/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 21
[Credential-related source omitted] - Access to secretsline 21
[Credential-related source omitted] - Network accessline 40
def get_json(url, timeout=30): req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 41
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
with urllib.request.urlopen(req, timeout=timeout) as resp: - Network accessline 46
def post_json(url, payload, timeout=30): body = json.dumps(payload).encode("utf-8") req = urllib.request.Request( - Network accessline 47
body = json.dumps(payload).encode("utf-8") req = urllib.request.Request( url, data=body, headers={"Content-Type": "application/json"} - Network accessline 48
req = urllib.request.Request( - Network accessline 49
url, data=body, headers={"Content-Type": "application/json"} ) with urllib.request.urlopen(req, timeout=timeout) as resp: - Network accessline 50
) with urllib.request.urlopen(req, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 51
with urllib.request.urlopen(req, timeout=timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 73
curl -s -X POST "$TASK_ENV_BASE_URL/api/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/reference/workbench-api.md
- Network accessline 37
curl -s -X POST "$BASE/api/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 9
python3 fetch_deal.py PRJ_EXAMPLE --summary # compact triage view python3 fetch_deal.py PRJ_EXAMPLE --out bundle.json python3 fetch_deal.py PRJ_EXAMPLE --base-url http://host:9020/ - Network accessline 11
python3 fetch_deal.py PRJ_EXAMPLE --base-url http://host:9020/ - Network accessline 13
Base URL resolution order: --base-url, $TASK_ENV_BASE_URL, $GDPEVO_ENV_BASE_URL, then http://task-env:9020/. - Network accessline 14
then http://task-env:9020/. - Network accessline 24
DEFAULT_BASE = "http://task-env:9020/" - Access to secretsline 45
[Credential-related source omitted] - Access to secretsline 46
[Credential-related source omitted] - Network accessline 53
url = base + path try: with urllib.request.urlopen(url, timeout=timeout) as resp: - Network accessline 54
try: with urllib.request.urlopen(url, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 55
with urllib.request.urlopen(url, timeout=timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_03/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 50
[Credential-related source omitted] - Network accessline 63
def get(base, path): try: with urllib.request.urlopen(base + path, timeout=30) as resp: - Network accessline 64
try: with urllib.request.urlopen(base + path, timeout=30) as resp: return json.loads(resp.read().decode()) - Network accessline 65
with urllib.request.urlopen(base + path, timeout=30) as resp: - Network accessline 73
[Credential-related source omitted] - Network accessline 74
[Credential-related source omitted] - Network accessline 75
req = urllib.request.Request( - Network accessline 77
headers={"Content-Type": "application/json"}) try: with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 78
try: with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode()) - Network accessline 79
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
def get(url): try: with urllib.request.urlopen(url, timeout=30) as r: - Network accessline 19
try: with urllib.request.urlopen(url, timeout=30) as r: return json.loads(r.read().decode()) - Network accessline 20
with urllib.request.urlopen(url, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_02/scripts/check_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/reflect-3/reflect-3_attempt_03/scripts/pull_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 29
def get(url): try: with urllib.request.urlopen(url, timeout=TIMEOUT) as r: - Network accessline 30
try: with urllib.request.urlopen(url, timeout=TIMEOUT) as r: return json.loads(r.read().decode()) - Network accessline 31
with urllib.request.urlopen(url, timeout=TIMEOUT) as r: - Network accessline 56
def fetch(base, deal_id): - Network accessline 189
[Credential-related source omitted] - Network accessline 190
sys.exit("--sql requires --sql-token") req = urllib.request.Request( args.base_url.rstrip("/") + "/api/query", - Network accessline 191
req = urllib.request.Request( - Network accessline 193
[Credential-related source omitted] - Network accessline 194
headers={"Content-Type": "application/json"}) with urllib.request.urlopen(req, timeout=TIMEOUT) as r: print(r.read().decode()) - Network accessline 195
with urllib.request.urlopen(req, timeout=TIMEOUT) as r: - Network accessline 199
bundle = fetch(args.base_url, args.deal_id)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_01/references/workbench-api.md
- Network accessline 47
curl -s -X POST "$BASE/api/query" -H 'Content-Type: application/json' \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_01/scripts/fetch_deal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 39
[Credential-related source omitted] - Network accessline 46
url = f"{base}{path}" try: with urllib.request.urlopen(url, timeout=timeout) as resp: - Network accessline 47
try: with urllib.request.urlopen(url, timeout=timeout) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 48
with urllib.request.urlopen(url, timeout=timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_01/scripts/units.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_02/SKILL.md
- Network accessline 44
curl -s -o /dev/null -w '%{http_code}\n' "$BASE/api/deals"
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_02/scripts/extract_contract.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_02/scripts/wb.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 21
[Credential-related source omitted] - Access to secretsline 21
[Credential-related source omitted] - Access to secretsline 22
[Credential-related source omitted] - Network accessline 46
def get(base, path): req = urllib.request.Request(_url(base, path), headers={"Accept": "application/json"}) - Network accessline 47
req = urllib.request.Request(_url(base, path), headers={"Accept": "application/json"}) - Network accessline 48
with urllib.request.urlopen(req, timeout=TIMEOUT) as r: - Network accessline 52
[Credential-related source omitted] - Network accessline 53
[Credential-related source omitted] - Network accessline 54
req = urllib.request.Request( - Network accessline 56
headers={"Content-Type": "application/json"}, ) with urllib.request.urlopen(req, timeout=TIMEOUT) as r: - Network accessline 57
) with urllib.request.urlopen(req, timeout=TIMEOUT) as r: return json.loads(r.read().decode()) - Network accessline 58
with urllib.request.urlopen(req, timeout=TIMEOUT) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_03/SKILL.md
- Network accessline 39
export TASK_ENV_BASE_URL=http://task-env:9020 # or the value the prompt/env file gives
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_03/references/workbench_api.md
- Network accessline 43
curl -s -X POST "$TASK_ENV_BASE_URL/api/query" \
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_03/scripts/fetch_deal.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 3
# including the playbook/policy that the deal record actually points to. # # Usage: TASK_ENV_BASE_URL=http://host:port ./fetch_deal.sh PRJ_XXXX > deal.json - Network accessline 4
# # Usage: TASK_ENV_BASE_URL=http://host:port ./fetch_deal.sh PRJ_XXXX > deal.json # - Network accessline 5
# Usage: TASK_ENV_BASE_URL=http://host:port ./fetch_deal.sh PRJ_XXXX > deal.json - Network accessline 18
BASE="${TASK_ENV_BASE_URL:-${GDPEVO_ENV_BASE_URL:-http://task-env:9020}}" - Access to secretsline 20
[Credential-related source omitted] - Network accessline 24
get() { curl -sS -m 20 "$BASE$1" 2>/dev/null; } - Network accessline 26
sql() { curl -sS -m 20 -X POST "$BASE/api/query" \ - Network accessline 27
curl -sS -m 20 -X POST "$BASE/api/query" \ - Dangerous shell operationline 68
trap 'rm -rf "$tmp"' EXIT
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_020/self/self_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_01/scripts/hub_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 48
if params: url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers=self._headers()) - Network accessline 49
url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers=self._headers()) with urllib.request.urlopen(req, timeout=60) as r: - Network accessline 50
req = urllib.request.Request(url, headers=self._headers()) - Network accessline 51
with urllib.request.urlopen(req, timeout=60) as r: - Network accessline 67
def sql(self, query): req = urllib.request.Request( - Network accessline 68
req = urllib.request.Request( - Network accessline 72
method="POST", ) with urllib.request.urlopen(req, timeout=120) as r: - Network accessline 73
) with urllib.request.urlopen(req, timeout=120) as r: return json.loads(r.read().decode()) - Network accessline 74
with urllib.request.urlopen(req, timeout=120) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_02/scripts/hub_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 73
headers["Content-Type"] = "application/json" data = json.dumps(data).encode() req = urllib.request.Request(self.base + path, data=data, headers=headers) - Network accessline 74
data = json.dumps(data).encode() req = urllib.request.Request(self.base + path, data=data, headers=headers) try: - Network accessline 75
req = urllib.request.Request(self.base + path, data=data, headers=headers) - Network accessline 76
try: with urllib.request.urlopen(req, timeout=self.timeout) as r: return r.status, json.loads(r.read().decode()) - Network accessline 77
with urllib.request.urlopen(req, timeout=self.timeout) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_03/references/environment.md
- Network accessline 3
## Credentials (never hard-code — read them at runtime) The task ships an `environment_access.md` in the working directory. It contains: - `GDPEVO_ENV_BASE_URL=<base url>` (e.g. `http://task-env:9021/`) - Network accessline 4
[Credential-related source omitted] - Network accessline 5
- `GDPEVO_ENV_BASE_URL=<base url>` (e.g. `http://task-env:9021/`)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_03/scripts/hub_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 79
def _open(self, req): try: with urllib.request.urlopen(req, timeout=self.timeout) as r: - Network accessline 80
try: with urllib.request.urlopen(req, timeout=self.timeout) as r: return r.status, json.loads(r.read().decode("utf-8", "replace")) - Network accessline 81
with urllib.request.urlopen(req, timeout=self.timeout) as r: - Network accessline 87
def get(self, path): url = self.base + (path if path.startswith("/") else "/" + path) req = urllib.request.Request(url, headers=self._auth, method="GET") - Network accessline 88
url = self.base + (path if path.startswith("/") else "/" + path) req = urllib.request.Request(url, headers=self._auth, method="GET") return self._open(req)[1] - Network accessline 89
req = urllib.request.Request(url, headers=self._auth, method="GET") - Network accessline 107
url = self.base + "/api/query" data = json.dumps({"query": sql}).encode() req = urllib.request.Request( - Network accessline 108
data = json.dumps({"query": sql}).encode() req = urllib.request.Request( url, headers={**self._auth, "Content-Type": "application/json"}, - Network accessline 109
req = urllib.request.Request(
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_01/reference/reconciliation_toolkit.py
- Executable scriptline 1
Executable file
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_02/scripts/hub_toolkit.py
- Executable scriptline 1
Executable file - Network accessline 72
if params: url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers=self._headers(), method="GET") - Network accessline 73
url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers=self._headers(), method="GET") with urllib.request.urlopen(req, timeout=60) as r: - Network accessline 74
req = urllib.request.Request(url, headers=self._headers(), method="GET") - Network accessline 75
with urllib.request.urlopen(req, timeout=60) as r: - Network accessline 88
while True: body = json.dumps({"query": sql + offset_clause}).encode() req = urllib.request.Request( - Network accessline 89
body = json.dumps({"query": sql + offset_clause}).encode() req = urllib.request.Request( self.base + path, data=body, method="POST", - Network accessline 90
req = urllib.request.Request( - Network accessline 91
self.base + path, data=body, method="POST", headers={**self._headers(), "Content-Type": "application/json"}) with urllib.request.urlopen(req, timeout=120) as r: - Network accessline 92
headers={**self._headers(), "Content-Type": "application/json"}) with urllib.request.urlopen(req, timeout=120) as r: resp = json.loads(r.read()) - Network accessline 93
with urllib.request.urlopen(req, timeout=120) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/reflect-3/reflect-3_attempt_03/references/hub_client.py
- Executable scriptline 1
Executable file - Access to secretsline 34
[Credential-related source omitted] - Access to secretsline 36
[Credential-related source omitted] - Network accessline 50
if data is not None: headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 51
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 52
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 53
try: with urllib.request.urlopen(req, timeout=60) as r: raw = r.read().decode() - Network accessline 54
with urllib.request.urlopen(req, timeout=60) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_021/self/self_attempt_01/scripts/asteria_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 45
The file format seen across tasks: GDPEVO_ENV_BASE_URL=http://task-env:9021/ - Network accessline 46
GDPEVO_ENV_BASE_URL=http://task-env:9021/ - Network accessline 97
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 98
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, method=method, headers=headers) try: - Network accessline 99
req = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 100
try: with urllib.request.urlopen(req, timeout=self.timeout) as resp: raw = resp.read().decode("utf-8") - Network accessline 101
with urllib.request.urlopen(req, timeout=self.timeout) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 58
Use the bundled helper to avoid re-writing curl each time:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 45
[Credential-related source omitted] - Network accessline 84
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 85
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 86
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 87
try: with urllib.request.urlopen(req, timeout=60) as resp: raw = resp.read().decode("utf-8") - Network accessline 88
with urllib.request.urlopen(req, timeout=60) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_03/scripts/atlas_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 83
last = None for attempt in range(retries): req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 84
for attempt in range(retries): req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 85
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 86
try: with urllib.request.urlopen(req, timeout=self.timeout) as resp: return json.loads(resp.read().decode()) - Network accessline 87
with urllib.request.urlopen(req, timeout=self.timeout) as resp: - Access to secretsline 159
[Credential-related source omitted]
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_01/scripts/atlas_query.sh
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_022/self/self_attempt_02/reference/patterns.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_01/references/methods.md
- Dangerous shell operationline 221
`total`: `magnitude = SE_b · sqrt(df · rY · rM / (1 - rM))`. For each declared bias direction
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_02/references/methods.md
- Dangerous shell operationline 189
`magnitude = SE_b · sqrt(df · rY · rM / (1 − rM))`. For each declared direction
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_03/references/portal.md
- Network accessline 9
The base URL is provided out-of-band, not in the task payload. Read it from `environment_access.md` (variable `GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`). Task prompts refer to it as `<TASK_ENV_BASE_URL>`. - Network accessline 10
`environment_access.md` (variable `GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`). Task prompts refer to it as `<TASK_ENV_BASE_URL>`. Confirm connectivity with `GET /` before doing analysis. - Network accessline 11
`http://task-env:9023/`). Task prompts refer to it as `<TASK_ENV_BASE_URL>`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_01/assets/portal.py
- Executable scriptline 1
Executable file - Access to secretsline 44
[Credential-related source omitted] - Network accessline 54
def _get(self, path: str) -> str: url = self.base + path with urllib.request.urlopen(url, timeout=self.timeout) as r: - Network accessline 55
url = self.base + path with urllib.request.urlopen(url, timeout=self.timeout) as r: return r.read().decode("utf-8", "replace") - Network accessline 56
with urllib.request.urlopen(url, timeout=self.timeout) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_02/references/portal_toolkit.py
- Executable scriptline 1
Executable file - Network accessline 30
q.update({k: v for k, v in filters.items() if v is not None}) url = self.base + "/download?" + urllib.parse.urlencode(q) with urllib.request.urlopen(url, timeout=120) as r: - Network accessline 31
url = self.base + "/download?" + urllib.parse.urlencode(q) with urllib.request.urlopen(url, timeout=120) as r: text = r.read().decode() - Network accessline 32
with urllib.request.urlopen(url, timeout=120) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/reflect-3/reflect-3_attempt_03/scripts/audit_toolkit.py
- Executable scriptline 1
Executable file
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_02/SKILL.md
- Network accessline 39
- `references/audit_modules.md` — the recurring module family and the exact reproducibility contract each module must satisfy. - `scripts/portal_client.py` — a GET-only portal CSV client (`Portal().fetch(...)`). - Network accessline 41
- `scripts/portal_client.py` — a GET-only portal CSV client (`Portal().fetch(...)`).
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_02/references/portal.md
- Network accessline 3
The Public Health Observatory (PHO) portal is the **only** authoritative evidence source for these tasks. Base URL comes from `environment_access.md` (`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`). All access is GET-only. - Network accessline 5
(`GDPEVO_ENV_BASE_URL`, e.g. `http://task-env:9023/`). All access is GET-only. - Network accessline 16
- `GET /catalog` — schema + row counts + filterable columns + measure dictionary. - `GET /methodology` and `GET /methodology?doc=<id>` — audit conventions (below). - `scripts/portal_client.py` wraps all of this: `Portal().fetch("state_healt - Network accessline 18
- `scripts/portal_client.py` wraps all of this: `Portal().fetch("state_health", year="2023")`.
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_02/scripts/portal_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 16
from portal_client import Portal p = Portal() # auto-discovers base URL rows = p.fetch("state_health", # list[dict], all columns as strings - Network accessline 17
p = Portal() # auto-discovers base URL rows = p.fetch("state_health", # list[dict], all columns as strings measure_id="life_expectancy", year="2023") - Network accessline 18
rows = p.fetch("state_health", # list[dict], all columns as strings - Network accessline 20
states = p.fetch("states") # geography reference (fips/abbr/region/division) - Access to secretsline 58
[Credential-related source omitted] - Network accessline 82
def _get(self, path: str) -> bytes: url = self.base_url + path with urllib.request.urlopen(url, timeout=self.timeout) as resp: - Network accessline 83
url = self.base_url + path with urllib.request.urlopen(url, timeout=self.timeout) as resp: return resp.read() - Network accessline 84
with urllib.request.urlopen(url, timeout=self.timeout) as resp: - Network accessline 93
def fetch(self, dataset: str, **filters: str) -> list[dict]: - Network accessline 134
dataset = rest[0] filters = dict(kv.split("=", 1) for kv in rest[1:] if "=" in kv) rows = p.fetch(dataset, **filters) - Network accessline 135
filters = dict(kv.split("=", 1) for kv in rest[1:] if "=" in kv) rows = p.fetch(dataset, **filters) w = csv.writer(sys.stdout) - Network accessline 136
rows = p.fetch(dataset, **filters)
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_03/SKILL.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_03/references/portal_reference.md
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_023/self/self_attempt_03/scripts/portal_fetch.py
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_01/scripts/portfolio_analysis.py
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_01/skill/scripts/portfolio_analysis.py
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_02/scripts/portfolio_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
def _request(url, data=None, headers=None): req = urllib.request.Request(url, data=data, headers=headers or {}, - Network accessline 42
req = urllib.request.Request(url, data=data, headers=headers or {}, - Network accessline 43
method="POST" if data else "GET") with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 44
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/fewshot/fewshot_attempt_03/portfolio_toolkit.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 44
[Credential-related source omitted] - Network accessline 45
req = urllib.request.Request(base + path) - Network accessline 46
[Credential-related source omitted] - Network accessline 47
req.add_header("X-Env-Token", token) with urllib.request.urlopen(req, timeout=30) as r: return json.load(r) - Network accessline 48
with urllib.request.urlopen(req, timeout=30) as r: - Network accessline 70
body = json.dumps({"sql": sql}).encode() req = urllib.request.Request(base + "/api/query", data=body, method="POST") - Network accessline 71
req = urllib.request.Request(base + "/api/query", data=body, method="POST") - Network accessline 73
[Credential-related source omitted] - Network accessline 74
req.add_header("X-Env-Token", token) with urllib.request.urlopen(req, timeout=30) as r: return json.load(r) - Network accessline 75
with urllib.request.urlopen(req, timeout=30) as r:
experiments/claude_code_opus_4_8_xhigh/reports/skills/task_group_024/self/self_attempt_03/scripts/query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 29
def get(base, path): with urllib.request.urlopen(base + path, timeout=15) as r: - Network accessline 30
with urllib.request.urlopen(base + path, timeout=15) as r: - Network accessline 36
sys.exit("POST /api/query needs a token, but none was found in environment_access.md") body = json.dumps({"sql": statement}).encode() req = urllib.request.Request( - Network accessline 37
body = json.dumps({"sql": statement}).encode() req = urllib.request.Request( base + "/api/query", - Network accessline 38
req = urllib.request.Request( - Network accessline 41
[Credential-related source omitted] - Network accessline 42
) with urllib.request.urlopen(req, timeout=15) as r: return json.load(r) - Network accessline 43
with urllib.request.urlopen(req, timeout=15) as r:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_002/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill/SKILL.md
- Network accessline 12
1. Resolve the API base URL from the task prompt or runner environment. Prompts may name it `API_BASE_URL`, `BASE_URL`, or show `http://127.0.0.1:<PORT>`.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01/SKILL.md
- Network accessline 10
1. Read the user prompt and the provided answer template first. Treat the template as authoritative for top-level keys, nested fields, nullability, enum vocabulary, and whether policy-code blocks must be returned. 2. Use only the account ID - Network accessline 11
2. Use only the account IDs, date ranges, months, regions, quarters, events, and due dates requested by the prompt. Do not add extra accounts unless the prompt asks for an all-region/all-customer summary. 3. Query the ApexCloud Retention Op - Network accessline 12
3. Query the ApexCloud Retention Operations API using the base URL from the current prompt or harness. In this eval family, a shared proxy may be provided at `http://127.0.0.1:8066` even when prompt text references the original local servic
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/demonstration_skill/demonstration_skill_attempt_02/apexcloud-retention-analytics/SKILL.md
- Network accessline 20
Use the API base URL explicitly allowed by the current task or controlling instruction. In evaluation harnesses, a proxy base such as `http://127.0.0.1:8066` can supersede a historical port embedded in sample prompts.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_01/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_02/SKILL.md
- Network accessline 12
1. Read the prompt and the supplied `input/payloads/answer_template.json` before calculating anything. The prompt sometimes omits fields that the template requires, especially `policy_codes`. 2. Use the remote API entrypoint given by the ta - Network accessline 13
2. Use the remote API entrypoint given by the task environment. If the prompt says `http://127.0.0.1:8074` but the evaluation workspace provides a remote base URL, replace only the host/base with the provided remote base and keep the same p
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_004/self/self_attempt_03/SKILL.md
- Network accessline 12
1. Read the user prompt and any `input/payloads/answer_template.json` before querying data. The template is the schema source of truth when it exists, including fields such as `policy_codes` that may not be repeated in the prose. 2. Use the - Network accessline 13
2. Use the environment entrypoint from `environment_access.md` when present. If a prompt says `http://127.0.0.1:8074` but the evaluation supplies a remote `GDPEVO_ENV_BASE_URL`, replace the local URL with the remote base URL and keep the sa
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_005/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02/SKILL.md
- Network accessline 23
For localhost calls in a proxied shell, use `curl --noproxy '*'`.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02/SKILL.md
- Network accessline 25
curl --noproxy '*' -sS "$BASE_URL/"
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_006/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03/SKILL.md
- Network accessline 12
1. Read the prompt, every file under the task's `input/payloads/`, and the answer template. 2. Query the ProcureOps service. Use the base URL from the prompt/runner, commonly `<environment_base_url>`. If localhost curl uses a proxy, call it - Network accessline 13
2. Query the ProcureOps service. Use the base URL from the prompt/runner, commonly `<environment_base_url>`. If localhost curl uses a proxy, call it as `curl --noproxy '*' -sS "$BASE/<endpoint>"`.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_01/demonstration-skill-attempt-01/SKILL.md
- Network accessline 25
http://127.0.0.1:8100 - Network accessline 43
Some collection endpoints may return more records than needed. Filter locally to the memo's order IDs, wave, BOM IDs, supplier IDs, or date window. Use `curl --noproxy '*'` or a Python HTTP client with proxy bypass if the environment has pr
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_02/demonstration-skill-attempt-02/SKILL.md
- Network accessline 12
- Use the public API, not environment files. In this benchmark the shared base URL is `http://127.0.0.1:8100`; if the runner or prompt gives a different public base URL, use that. - If proxy settings interfere, call the API with `curl --nop - Network accessline 13
- If proxy settings interfere, call the API with `curl --noproxy '*' -sS '<url>'` or configure the HTTP client to bypass proxies for `127.0.0.1`.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_007/demonstration_skill/demonstration_skill_attempt_03/demonstration-skill-attempt-03/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_01/task-group-010-input-output/SKILL.md
- Network accessline 1
--- name: task-group-010-fewshot-attempt-01 description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks that require reading local request/template JSON, using the public Asteria API at http://127.0.0.1:8036 a - Network accessline 2
name: task-group-010-fewshot-attempt-01 description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks that require reading local request/template JSON, using the public Asteria API at http://127.0.0.1:8036 as th - Network accessline 3
description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks that require reading local request/template JSON, using the public Asteria API at http://127.0.0.1:8036 as the current book of record, computing port - Network accessline 12
1. Read the local `input/prompt.txt`, request payload, and `input/payloads/answer_template.json`. 2. Treat the public API at `http://127.0.0.1:8036` as authoritative for current portfolio, policy, instrument, issuer, index, and signal recor - Network accessline 13
2. Treat the public API at `http://127.0.0.1:8036` as authoritative for current portfolio, policy, instrument, issuer, index, and signal records.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_02/asteria-portfolio-risk-json/SKILL.md
- Network accessline 10
1. Read the local prompt, request payload, and `answer_template.json` first. Treat the template as the output contract for required fields, ordering, allowed enum values, and numeric precision. 2. Use only the public API at `http://127.0.0. - Network accessline 11
2. Use only the public API at `http://127.0.0.1:8036` as the current book of record. Local payloads often include stale snapshots or preference memos; use them for requested IDs, ticket counts, windows, focus lists, and committee context, n
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/demonstration_skill/demonstration_skill_attempt_03/task-group-010-asteria-portfolio-risk/SKILL.md
- Network accessline 1
--- name: task-group-010-fewshot-attempt-03 description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks. Use when producing JSON answers from local request payloads and the public Asteria API at http://127.0.0 - Network accessline 2
name: task-group-010-fewshot-attempt-03 description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks. Use when producing JSON answers from local request payloads and the public Asteria API at http://127.0.0.1:8 - Network accessline 3
description: Solve PanofyBench task_group_010 Asteria Investment Office input-output tasks. Use when producing JSON answers from local request payloads and the public Asteria API at http://127.0.0.1:8036 for credit trade packages, fixed-inc - Network accessline 8
Use only the local task input files and the public API at `http://127.0.0.1:8036`. Treat the API as the current book of record; local payload snapshots, desk notes, and stale shortlists are request context only unless the template explicitl
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_010/self/self_attempt_03/SKILL.md
- Network accessline 7
1. Read the task prompt and `input/payloads/*` files first. Treat local memos, stale snapshots, and preference packets as context, not authoritative holdings or marks. 2. Use the environment access file for the API base URL. Quote URLs that - Network accessline 8
2. Use the environment access file for the API base URL. Quote URLs that contain query strings, for example `curl -s '.../api/instruments/bonds?candidate=true'`.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_01/SKILL.md
- Network accessline 22
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/health' - Network accessline 23
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/employees?q=<employee_id_or_name>' - Network accessline 24
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/cases/<case_id>' - Network accessline 25
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/recruitment?q=<opening_id>' - Network accessline 26
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/payroll-ledgers?q=<employee_id_or_name>' - Network accessline 27
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/documents?q=<case_or_employee_or_document_id>' - Network accessline 28
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/messages?q=<case_or_message_or_candidate_id>' - Network accessline 29
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/audit?q=<case_or_employee_or_opening_id>' - Network accessline 30
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/audit/<audit_id>' - Network accessline 31
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/policies?q=<policy_keyword>' - Network accessline 32
curl --noproxy '*' -sS 'http://127.0.0.1:<port>/api/policies/<policy_id>'
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/demonstration_skill/demonstration_skill_attempt_02/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_012/self/self_attempt_01/SKILL.md
- Network accessline 7
1. Read the prompt and `input/payloads/answer_template.json` first. Treat the template as the output contract: exact keys, exact enum spellings, booleans as booleans, numbers as numbers, and arrays as arrays of IDs unless the template says - Network accessline 8
2. Read `environment_access.md` and replace any prompt URL like `http://127.0.0.1:<port>/` with the provided remote base URL. Use the credentials from that file if opening the UI.
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 25
curl -sS -X POST "$TASK_ENV_BASE_URL/query" \
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/fewshot_attempt_03/SKILL.md
- Network accessline 19
BASE="${TASK_ENV_BASE_URL:-http://task-env:9015}" curl -fsS "$BASE/api/patients/<patient_id>" | jq . - Network accessline 20
curl -fsS "$BASE/api/patients/<patient_id>" | jq . - Network accessline 21
curl -fsS "$BASE/api/referrals" | jq . - Network accessline 55
curl -fsS "$BASE/api/patients/<patient_id>/service-requests" | jq .
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_02/SKILL.md
- Network accessline 19
BASE="<hub base url>" KEY="<api key if supplied>" curl -sS "$BASE/api/schema" - Network accessline 20
KEY="<api key if supplied>" curl -sS "$BASE/api/schema" curl -sS "$BASE/api/matters?matter_id=<matter_id>" - Network accessline 21
curl -sS "$BASE/api/schema" - Network accessline 22
curl -sS "$BASE/api/matters?matter_id=<matter_id>" - Network accessline 23
curl -sS "$BASE/api/subpoena-categories?matter_id=<matter_id>" - Network accessline 24
curl -sS "$BASE/api/productions?matter_id=<matter_id>" - Network accessline 25
curl -sS "$BASE/api/custodian-sources?matter_id=<matter_id>" - Network accessline 26
curl -sS "$BASE/api/documents/search?matter_id=<matter_id>" - Network accessline 27
curl -sS "$BASE/api/privilege-log?matter_id=<matter_id>" - Network accessline 28
curl -sS "$BASE/api/qc-findings?matter_id=<matter_id>" - Network accessline 29
curl -sS "$BASE/api/retention-events?matter_id=<matter_id>" - Network accessline 30
curl -sS "$BASE/api/remediation-actions?matter_id=<matter_id>" - Network accessline 36
[Credential-related source omitted]
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_02/scripts/fetch_hub_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 27
[Credential-related source omitted] - Network accessline 28
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) url = f"{url}?{urllib.parse.urlencode({'matter_id': matter_id})}" request = urllib.request.Request(url) - Network accessline 29
[Credential-related source omitted] - Network accessline 30
request = urllib.request.Request(url) - Access to secretsline 31
[Credential-related source omitted] - Network accessline 32
[Credential-related source omitted] - Access to secretsline 32
[Credential-related source omitted] - Network accessline 33
try: with urllib.request.urlopen(request, timeout=20) as response: return json.load(response) - Network accessline 34
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 53
[Credential-related source omitted]
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/fewshot_attempt_03/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/fewshot_attempt_02/scripts/licensing_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 91
[Credential-related source omitted] - Network accessline 92
headers["X-Task-Token"] = token req = urllib.request.Request(url, data=data, headers=headers) with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 93
req = urllib.request.Request(url, data=data, headers=headers) - Network accessline 94
with urllib.request.urlopen(req, timeout=20) as resp:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_02/scripts/collect_workbench_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 35
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 36
headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 37
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 38
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 61
def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description="Collect deal workbench data for one deal ID.") parser.add_argument("--base-url", required=True, help="Workbench base URL, for example http://task- - Network accessline 62
parser = argparse.ArgumentParser(description="Collect deal workbench data for one deal ID.") parser.add_argument("--base-url", required=True, help="Workbench base URL, for example http://task-env:9020/") parser.add_argument("--deal- - Network accessline 63
parser.add_argument("--base-url", required=True, help="Workbench base URL, for example http://task-env:9020/")
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/fewshot_attempt_03/scripts/fetch_deal_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 44
[Credential-related source omitted] - Network accessline 45
headers["Authorization"] = f"Bearer {token}" request = urllib.request.Request(url, headers=headers) with urllib.request.urlopen(request, timeout=10) as response: - Network accessline 46
request = urllib.request.Request(url, headers=headers) - Network accessline 47
with urllib.request.urlopen(request, timeout=10) as response: - Network accessline 51
[Credential-related source omitted] - Network accessline 52
[Credential-related source omitted] - Network accessline 53
request = urllib.request.Request( - Network accessline 57
method="POST", ) with urllib.request.urlopen(request, timeout=10) as response: - Network accessline 58
) with urllib.request.urlopen(request, timeout=10) as response: return json.load(response) - Network accessline 59
with urllib.request.urlopen(request, timeout=10) as response:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_01/scripts/query_asteria.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 38
[Credential-related source omitted] - Network accessline 39
payload = json.dumps({"query": args.query}).encode("utf-8") request = urllib.request.Request( f"{base_url}/api/query", - Network accessline 40
request = urllib.request.Request( - Network accessline 48
) try: with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 49
try: with urllib.request.urlopen(request, timeout=30) as response: data = json.loads(response.read().decode("utf-8")) - Network accessline 50
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_02/scripts/query_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
[Credential-related source omitted] - Network accessline 37
body = json.dumps({"query": sql}).encode("utf-8") req = urllib.request.Request( base_url + "api/query", - Network accessline 38
req = urllib.request.Request( - Network accessline 46
) try: with urllib.request.urlopen(req, timeout=60) as resp: - Network accessline 47
try: with urllib.request.urlopen(req, timeout=60) as resp: payload = json.load(resp) - Network accessline 48
with urllib.request.urlopen(req, timeout=60) as resp:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/fewshot_attempt_03/SKILL.md
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/self/self_attempt_01/SKILL.md
- Network accessline 20
curl -sS -X POST "$BASE/api/query" \
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_021/self/self_attempt_02/SKILL.md
- Network accessline 18
curl -sS -X POST "$GDPEVO_ENV_BASE_URL/api/query" \
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_01/SKILL.md
- Network accessline 20
curl -sS \
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_02/SKILL.md
- Access to secretsline 25
[Credential-related source omitted]
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_02/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 17
[Credential-related source omitted] - Access to secretsline 30
[Credential-related source omitted] - Access to secretsline 31
[Credential-related source omitted] - Access to secretsline 33
[Credential-related source omitted] - Access to secretsline 40
[Credential-related source omitted] - Network accessline 41
if body is not None: headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 42
headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, method=method, headers=headers) try: - Network accessline 43
request = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 44
try: with urllib.request.urlopen(request, timeout=args.timeout) as response: raw = response.read().decode("utf-8") - Network accessline 45
with urllib.request.urlopen(request, timeout=args.timeout) as response:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/fewshot_attempt_03/pho_audit_tools.py
- Executable scriptline 1
Executable file - Network accessline 64
query.update({k: v for k, v in params.items() if v is not None}) url = urllib.parse.urljoin(base, "download") + "?" + urllib.parse.urlencode(query) with urllib.request.urlopen(url) as resp: - Network accessline 65
url = urllib.parse.urljoin(base, "download") + "?" + urllib.parse.urlencode(query) with urllib.request.urlopen(url) as resp: text = resp.read().decode("utf-8") - Network accessline 66
with urllib.request.urlopen(url) as resp:
experiments/codex_gpt5_5_xhigh/reports/skills/task_group_024/self/self_attempt_01/SKILL.md
- Network accessline 21
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_01/scripts/normalize_contact.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_03/scripts/fetch_api.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_03/scripts/normalize.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_02/references/api-reference.md
- Network accessline 3
Base URL is supplied by the runner as `<TASK_ENV_BASE_URL>` (typically `http://task-env:9001`). All endpoints are read-only `GET`. No authentication is required.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_03/references/api.md
- Network accessline 3
Base URL is supplied by the runner as `<TASK_ENV_BASE_URL>` (typically `http://task-env:9001/`). All endpoints are GET. No authentication is required.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 46
Use `curl -s` against `{TASK_ENV_BASE_URL}/api/{collection}` or
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_01/references/api_endpoints.md
- Network accessline 3
Base URL: `<TASK_ENV_BASE_URL>` (provided by the runner, typically `http://task-env:9002`)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 15
curl -s "${BASE_URL}/api" - Network accessline 66
QUOTE=$(curl -s "${BASE_URL}/api/quotes/<quote_id>") - Network accessline 67
CUSTOMER_ID=$(echo "$QUOTE" | python3 -c "import sys,json; print(json.load(sys.stdin)['customer_id'])") PRODUCT_CODE=$(echo "$QUOTE" | python3 -c "import sys,json; print(json.load(sys.stdin)['primary_product_code'])") CUSTOMER=$(curl -s "${ - Network accessline 68
PRODUCT_CODE=$(echo "$QUOTE" | python3 -c "import sys,json; print(json.load(sys.stdin)['primary_product_code'])") CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") PRODUCT=$(curl -s "${BASE_URL}/api/products/${PRODUCT_CODE}") - Network accessline 69
CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") - Network accessline 70
PRODUCT=$(curl -s "${BASE_URL}/api/products/${PRODUCT_CODE}") - Network accessline 71
FREIGHT=$(curl -s "${BASE_URL}/api/freight-quotes") - Network accessline 72
POLICIES=$(curl -s "${BASE_URL}/api/policies") - Network accessline 118
RFQ=$(curl -s "${BASE_URL}/api/rfqs/<rfq_id>") - Network accessline 119
CUSTOMER_ID=$(echo "$RFQ" | python3 -c "import sys,json; print(json.load(sys.stdin)['customer_id'])") CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") POLICIES=$(curl -s "${BASE_URL}/api/policies") - Network accessline 120
CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") - Network accessline 121
POLICIES=$(curl -s "${BASE_URL}/api/policies") - Network accessline 129
PRODUCT=$(curl -s "${BASE_URL}/api/products/<product_code>") - Network accessline 154
OPPORTUNITY=$(curl -s "${BASE_URL}/api/opportunities/<opportunity_id>") - Network accessline 155
CUSTOMER_ID=$(echo "$OPPORTUNITY" | python3 -c "import sys,json; print(json.load(sys.stdin)['customer_id'])") CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") INVOICES=$(curl -s "${BASE_URL}/api/invoices") - Network accessline 156
CUSTOMER=$(curl -s "${BASE_URL}/api/customers/${CUSTOMER_ID}") - Network accessline 157
INVOICES=$(curl -s "${BASE_URL}/api/invoices") - Network accessline 158
PAYMENTS=$(curl -s "${BASE_URL}/api/payments") - Network accessline 159
REVJOURNALS=$(curl -s "${BASE_URL}/api/revenue-journals") - Network accessline 160
EVENTS=$(curl -s "${BASE_URL}/api/events") - Network accessline 161
POLICIES=$(curl -s "${BASE_URL}/api/policies")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 18
BASE_URL="http://task-env:9002" - Network accessline 25
curl -s "$BASE_URL/health" - Network accessline 80
curl -s "$BASE_URL/api/quotes/<quote_id>" - Network accessline 81
curl -s "$BASE_URL/api/customers/<customer_id>" - Network accessline 82
curl -s "$BASE_URL/api/products/<product_code>" - Network accessline 83
curl -s "$BASE_URL/api/freight-quotes" # then filter by quote_id - Network accessline 84
curl -s "$BASE_URL/api/policies" - Network accessline 193
curl -s "$BASE_URL/api/rfqs/<rfq_id>" - Network accessline 194
curl -s "$BASE_URL/api/customers/<customer_id>" - Network accessline 195
curl -s "$BASE_URL/api/policies" - Network accessline 206
curl -s "$BASE_URL/api/products/<product_code>" - Network accessline 245
curl -s "$BASE_URL/api/opportunities/<opportunity_id>" - Network accessline 246
curl -s "$BASE_URL/api/customers/<customer_id>" - Network accessline 247
curl -s "$BASE_URL/api/invoices" - Network accessline 248
curl -s "$BASE_URL/api/payments" - Network accessline 249
curl -s "$BASE_URL/api/revenue-journals" - Network accessline 250
curl -s "$BASE_URL/api/events" - Network accessline 251
curl -s "$BASE_URL/api/vouchers/<voucher_code>" # if voucher named in prompt
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_03/reference/api.md
- Network accessline 3
Base URL: `$BASE_URL` as provided by the task runner (typically `http://task-env:9002`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 10
Resolve telecom support tasks against a shared REST API. Every task includes a `<TASK_ENV_BASE_URL>` placeholder that points to the running support-console instance (typically `http://task-env:9003/`). All data must come from API responses,
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_01/references/api_endpoints.md
- Network accessline 3
Base URL: `<TASK_ENV_BASE_URL>` (replace with actual URL, typically `http://task-env:9003`)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_02/scripts/query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 11
Examples: python3 scripts/query.py http://task-env:9003 /api/tickets/TCK-5107 - Network accessline 12
python3 scripts/query.py http://task-env:9003 /api/tickets/TCK-5107 - Network accessline 13
python3 scripts/query.py http://task-env:9003 /api/outages - Network accessline 14
python3 scripts/query.py http://task-env:9003 /api/catalog - Network accessline 25
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 26
try: with urllib.request.urlopen(req, timeout=10) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 27
with urllib.request.urlopen(req, timeout=10) as resp: - Network accessline 37
if len(sys.argv) < 3: print("Usage: query.py <base_url> <endpoint> [<resource_id>]") print("Example: query.py http://task-env:9003 /api/tickets/TCK-5107") - Network accessline 38
print("Usage: query.py <base_url> <endpoint> [<resource_id>]") print("Example: query.py http://task-env:9003 /api/tickets/TCK-5107") sys.exit(1) - Network accessline 39
print("Example: query.py http://task-env:9003 /api/tickets/TCK-5107")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_003/fewshot/deepagents/fewshot_attempt_03/scripts/example.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 439
The API supports many concurrent requests. When analyzing multiple accounts, fetch all per-account endpoints in parallel rather than sequentially. Use curl in backgrounded shell jobs or Python `concurrent.futures` to cut total - Network accessline 440
fetch all per-account endpoints in parallel rather than sequentially. Use curl in backgrounded shell jobs or Python `concurrent.futures` to cut total latency. - Network accessline 441
curl in backgrounded shell jobs or Python `concurrent.futures` to cut total
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_01/references/api_reference.md
- Network accessline 3
Base URL: `<TASK_ENV_BASE_URL>` or `http://task-env:9004/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 11
This skill provides the complete API catalog, controlled vocabularies, business rules, and reusable task patterns for the ApexCloud Retention Operations API. Use it when a task requires building structured retention analytics from the ApexC
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_01/references/api.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_02/SKILL.md
- Network accessline 14
The task prompt will give you the base URL (typically `http://task-env:9004`). Read the prompt carefully to extract: - Network accessline 29
The task prompt uses a placeholder like `<TASK_ENV_BASE_URL>`. Replace it with the actual URL before making requests (e.g. `http://task-env:9004`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_02/references/endpoints.md
- Network accessline 3
Base URL: the task prompt provides this (typically `http://task-env:9004`). All endpoints are GET requests with no authentication required.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 1
--- name: apexcloud-retention-ops description: Build structured retention operations outputs from the ApexCloud Retention Operations API (task-env:9004). Use when the task involves ApexCloud, retention operations, renewal risk queues, QBR m - Network accessline 2
name: apexcloud-retention-ops description: Build structured retention operations outputs from the ApexCloud Retention Operations API (task-env:9004). Use when the task involves ApexCloud, retention operations, renewal risk queues, QBR metri - Network accessline 3
description: Build structured retention operations outputs from the ApexCloud Retention Operations API (task-env:9004). Use when the task involves ApexCloud, retention operations, renewal risk queues, QBR metrics packets, receivables review - Network accessline 8
Build deterministic JSON outputs from the ApexCloud Retention Operations API. The API is at `<TASK_ENV_BASE_URL>` (typically `http://task-env:9004`) with no authentication.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03/references/api_reference.md
- Network accessline 3
Base URL: `<TASK_ENV_BASE_URL>` (typically `http://task-env:9004`). No authentication. All endpoints are GET.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_02/scripts/amortization.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 20
2. Read the answer template to understand the output schema and allowed values. 3. Query the API for each candidate ID across all relevant endpoints. Use `curl -s` with the runner-provided `<TASK_ENV_BASE_URL>`. Page through - Network accessline 21
3. Query the API for each candidate ID across all relevant endpoints. Use `curl -s` with the runner-provided `<TASK_ENV_BASE_URL>`. Page through results when needed with `?limit=100&offset=N`. - Network accessline 22
`curl -s` with the runner-provided `<TASK_ENV_BASE_URL>`. Page through
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/cc/fewshot_attempt_01/references/api_endpoints.md
- Network accessline 3
Base URL: provided as `TASK_ENV_BASE_URL` (typically `http://localhost:9006`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_02/scripts/fetch_all.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03/SKILL.md
- Network accessline 37
# Pull all records from endpoints likely relevant to the task curl -s <TASK_ENV_BASE_URL>/programs - Network accessline 38
curl -s <TASK_ENV_BASE_URL>/programs - Network accessline 39
curl -s <TASK_ENV_BASE_URL>/suppliers - Network accessline 40
curl -s <TASK_ENV_BASE_URL>/contracts - Network accessline 41
curl -s <TASK_ENV_BASE_URL>/purchase_orders - Network accessline 42
curl -s <TASK_ENV_BASE_URL>/receipts - Network accessline 43
curl -s <TASK_ENV_BASE_URL>/ap/invoices - Network accessline 44
curl -s <TASK_ENV_BASE_URL>/ap/payments - Network accessline 45
curl -s <TASK_ENV_BASE_URL>/purchase_requisitions - Network accessline 46
curl -s <TASK_ENV_BASE_URL>/approvals - Network accessline 47
curl -s <TASK_ENV_BASE_URL>/budget_snapshots - Network accessline 48
curl -s <TASK_ENV_BASE_URL>/vendor_risk_events - Network accessline 49
curl -s <TASK_ENV_BASE_URL>/items
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03/references/api_endpoints.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03/scripts/procureops_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 10
Examples: python3 procureops_client.py http://task-env:9006 programs - Network accessline 11
python3 procureops_client.py http://task-env:9006 programs - Network accessline 12
python3 procureops_client.py http://task-env:9006 purchase_orders --ids PO-AX17-4481,PO-AX17-4519 - Network accessline 13
python3 procureops_client.py http://task-env:9006 receipts --ids RCV-BLUE-14 - Network accessline 29
def fetch_endpoint(base_url: str, endpoint: str) -> dict: url = f"{base_url.rstrip('/')}/{endpoint}" with urllib.request.urlopen(url) as resp: - Network accessline 30
url = f"{base_url.rstrip('/')}/{endpoint}" with urllib.request.urlopen(url) as resp: return json.loads(resp.read().decode()) - Network accessline 31
with urllib.request.urlopen(url) as resp: - Network accessline 66
import argparse parser = argparse.ArgumentParser(description="ProcureOps API client") parser.add_argument("base_url", help="Base URL, e.g. http://task-env:9006") - Network accessline 67
parser = argparse.ArgumentParser(description="ProcureOps API client") parser.add_argument("base_url", help="Base URL, e.g. http://task-env:9006") parser.add_argument("endpoint", help=f"Endpoint: {', '.join(ENDPOINTS)}") - Network accessline 68
parser.add_argument("base_url", help="Base URL, e.g. http://task-env:9006")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_data.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
for ep in ENDPOINTS: url = f"{base}/{ep}" with urllib.request.urlopen(url) as resp: - Network accessline 38
url = f"{base}/{ep}" with urllib.request.urlopen(url) as resp: payload = json.loads(resp.read()) - Network accessline 39
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 192
Use `curl -s` with `jq` for API access. Common patterns: - Network accessline 195
# Fetch manifest first curl -s "<TASK_ENV_BASE_URL>/manifest" | jq . - Network accessline 196
curl -s "<TASK_ENV_BASE_URL>/manifest" | jq . - Network accessline 198
# Fetch a specific collection curl -s "<TASK_ENV_BASE_URL>/purchase_orders" | jq . - Network accessline 199
curl -s "<TASK_ENV_BASE_URL>/purchase_orders" | jq . - Network accessline 201
# Filter by program curl -s "<TASK_ENV_BASE_URL>/purchase_orders" | jq '[.[] | select(.program_id == "PRG-TARGET")]' - Network accessline 202
curl -s "<TASK_ENV_BASE_URL>/purchase_orders" | jq '[.[] | select(.program_id == "PRG-TARGET")]' - Network accessline 204
# Find receipts for a PO curl -s "<TASK_ENV_BASE_URL>/receipts" | jq '[.[] | select(.po_id == "PO-TARGET-XXXX")]' - Network accessline 205
curl -s "<TASK_ENV_BASE_URL>/receipts" | jq '[.[] | select(.po_id == "PO-TARGET-XXXX")]' - Network accessline 207
# Check for scheduled payments for an invoice through a date curl -s "<TASK_ENV_BASE_URL>/ap/payments" | jq '[.[] | select(.invoice_id == "AP-TARGET-XXXX" and .scheduled_date <= "YYYY-MM-DD")]' - Network accessline 208
curl -s "<TASK_ENV_BASE_URL>/ap/payments" | jq '[.[] | select(.invoice_id == "AP-TARGET-XXXX" and .scheduled_date <= "YYYY-MM-DD")]'
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_03/references/api.md
- Network accessline 86
orders = requests.get(f"{BASE}/orders").json() - Network accessline 87
products = {p["sku"]: p for p in requests.get(f"{BASE}/products").json()} - Network accessline 88
customers = {c["customer_id"]: c for c in requests.get(f"{BASE}/customers").json()} - Network accessline 89
inventory_list = requests.get(f"{BASE}/inventory").json() - Network accessline 90
warehouses = {w["warehouse_id"]: w for w in requests.get(f"{BASE}/warehouses").json()} - Network accessline 91
suppliers = {s["supplier_id"]: s for s in requests.get(f"{BASE}/suppliers").json()} - Network accessline 92
pos = requests.get(f"{BASE}/purchase_orders").json() - Network accessline 93
boms = {b["bom_id"]: b for b in requests.get(f"{BASE}/boms").json()} - Network accessline 94
incidents = requests.get(f"{BASE}/incidents").json()
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_03/scripts/northwind_helpers.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_01/scripts/northwind_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
[Credential-related source omitted] - Access to secretsline 31
[Credential-related source omitted] - Network accessline 37
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as resp: - Network accessline 38
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as resp: return json.loads(resp.read().decode()) - Network accessline 39
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_02/references/api.md
- Network accessline 5
Base URL is provided per-task as the `TASK_ENV_BASE_URL` placeholder (typically `http://task-env:9007`). No authentication required — all endpoints are open GET. All responses are `application/json`. List endpoints return full collections a
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/codex/fewshot_attempt_03/scripts/fetch_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 11
url = f"{base_url}/{endpoint}" with urllib.request.urlopen(url) as resp: - Network accessline 12
with urllib.request.urlopen(url) as resp: - Network accessline 19
url = f"{base_url}/{endpoint}/{entity_id}" try: with urllib.request.urlopen(url) as resp: - Network accessline 20
try: with urllib.request.urlopen(url) as resp: return json.loads(resp.read().decode()) - Network accessline 21
with urllib.request.urlopen(url) as resp: - Network accessline 50
if len(sys.argv) < 2: print("Usage: fetch_api.py <BASE_URL> [endpoint [entity_id]]", file=sys.stderr) print(" BASE_URL: e.g., http://task-env:9007", file=sys.stderr) - Network accessline 51
print("Usage: fetch_api.py <BASE_URL> [endpoint [entity_id]]", file=sys.stderr) print(" BASE_URL: e.g., http://task-env:9007", file=sys.stderr) print(" endpoint: products, inventory, warehouses, orders, customers,", file=s - Network accessline 52
print(" BASE_URL: e.g., http://task-env:9007", file=sys.stderr)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_01/references/api.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_02/scripts/erp_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 5
Load this module with curl/wget data already fetched and held in memory - Network accessline 24
qs = "&".join(f"{k}={v}" for k, v in params.items() if v is not None) url = f"{url}?{qs}" with urllib.request.urlopen(url) as resp: - Network accessline 25
url = f"{url}?{qs}" with urllib.request.urlopen(url) as resp: return json.loads(resp.read()) - Network accessline 26
with urllib.request.urlopen(url) as resp: - Network accessline 32
url = f"{base_url}{endpoint}/{id_val}" try: with urllib.request.urlopen(url) as resp: - Network accessline 33
try: with urllib.request.urlopen(url) as resp: return json.loads(resp.read()) - Network accessline 34
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 14
1. Read the task prompt and any payload files (memos, answer templates, request JSON). 2. Identify the base URL (`<TASK_ENV_BASE_URL>`, typically `http://task-env:9007`). 3. Fetch the relevant API collections — prefer full-list endpoints ov - Network accessline 15
2. Identify the base URL (`<TASK_ENV_BASE_URL>`, typically `http://task-env:9007`). 3. Fetch the relevant API collections — prefer full-list endpoints over individual lookups. 4. Cross-reference live API data against the task memo to build - Dangerous shell operationline 57
curl -s <BASE>/orders | python3 -m json.tool - Network accessline 57
curl -s <BASE>/orders | python3 -m json.tool - Dangerous shell operationline 58
curl -s <BASE>/products | python3 -m json.tool - Network accessline 58
curl -s <BASE>/products | python3 -m json.tool - Dangerous shell operationline 59
curl -s <BASE>/inventory | python3 -m json.tool - Network accessline 59
curl -s <BASE>/inventory | python3 -m json.tool - Dangerous shell operationline 60
curl -s <BASE>/customers | python3 -m json.tool - Network accessline 60
curl -s <BASE>/customers | python3 -m json.tool - Dangerous shell operationline 61
curl -s <BASE>/suppliers | python3 -m json.tool - Network accessline 61
curl -s <BASE>/suppliers | python3 -m json.tool - Dangerous shell operationline 62
curl -s <BASE>/warehouses | python3 -m json.tool - Network accessline 62
curl -s <BASE>/warehouses | python3 -m json.tool - Dangerous shell operationline 63
curl -s <BASE>/incidents | python3 -m json.tool - Network accessline 63
curl -s <BASE>/incidents | python3 -m json.tool - Dangerous shell operationline 64
curl -s <BASE>/purchase_orders | python3 -m json.tool - Network accessline 64
curl -s <BASE>/purchase_orders | python3 -m json.tool - Dangerous shell operationline 65
curl -s <BASE>/boms | python3 -m json.tool - Network accessline 65
curl -s <BASE>/boms | python3 -m json.tool
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_03/references/api_reference.md
- Network accessline 7
The base URL is provided by the task runner as `<TASK_ENV_BASE_URL>`, typically `http://task-env:9007`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_03/scripts/compute_rmd.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_01/references/api_reference.md
- Network accessline 6
All responses are JSON. Use curl -s and pipe through Python for filtering when needed.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_02/references/api_reference.md
- Network accessline 3
Base URL is `$API_BASE`, typically `http://task-env:9008/`. All responses are JSON. Use `curl -s` with `python3 -m json.tool` for pretty-printing, or parse - Network accessline 4
JSON. Use `curl -s` with `python3 -m json.tool` for pretty-printing, or parse - Network accessline 139
def fetch_json(url): with urllib.request.urlopen(url) as resp: - Network accessline 140
with urllib.request.urlopen(url) as resp: - Network accessline 143
base = "http://task-env:9008"
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_03/references/api_endpoints.md
- Network accessline 3
Base URL: `{API_BASE}` (supplied by the harness, usually `http://task-env:9008/`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_01/scripts/crummey_timeline.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_01/scripts/grat_crat.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_01/scripts/roth_projection.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_02/references/api_endpoints.md
- Network accessline 3
Base URL is supplied by the harness as `API_BASE` (e.g. `http://task-env:9008/`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_03/scripts/verify.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_01/compute.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 22
All endpoints are served from the base URL supplied as `API_BASE` in the harness (typically `http://task-env:9008/`). Use `GET` for all requests. No authentication is required.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 12
The task environment provides a private wealth advisory API. Locate the base URL from the harness-supplied `API_BASE` environment variable; if unset, default to `http://task-env:9008/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_03/calc.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 15
- A request-memo file naming a client ID and engagement type - An answer template JSON defining the required output shape - The advisory API at `http://task-env:9008` - Network accessline 16
- An answer template JSON defining the required output shape - The advisory API at `http://task-env:9008` - Client records that may conflict across multiple source systems - Network accessline 17
- The advisory API at `http://task-env:9008` - Network accessline 20
The API base URL is supplied by the task harness, usually exposed as `API_BASE` after the environment starts. If not set, default to `http://task-env:9008`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_01/references/api-reference.md
- Network accessline 3
Base URL: `http://task-env:9008` (or `API_BASE` if set by the harness). All endpoints return JSON.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_02/scripts/compute.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_03/references/api-endpoints.md
- Network accessline 3
Base URL: `API_BASE` (supplied by harness, e.g. `http://task-env:9008`)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_01/SKILL.md
- Network accessline 37
the specific API endpoints available for this task. Cross-reference with `environment_access.md` in the workspace root to get the actual base URL (e.g. `http://task-env:9009/`). If `environment_access.md` is absent, use - Network accessline 38
`environment_access.md` in the workspace root to get the actual base URL (e.g. `http://task-env:9009/`). If `environment_access.md` is absent, use the `base_url` field directly. - Network accessline 39
(e.g. `http://task-env:9009/`). If `environment_access.md` is absent, use - Network accessline 61
curl -s http://task-env:9009/api/finance/branches - Network accessline 62
curl -s http://task-env:9009/api/finance/records - Network accessline 166
For any computation beyond simple aggregation, especially compensation and payroll, write a small Python script. It keeps the work auditable and reduces arithmetic errors. Use the API data as input (pipe from curl or save to a temp - Network accessline 167
payroll, write a small Python script. It keeps the work auditable and reduces arithmetic errors. Use the API data as input (pipe from curl or save to a temp file) and print the final answer JSON to stdout. - Network accessline 168
arithmetic errors. Use the API data as input (pipe from curl or save to a temp
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_02/SKILL.md
- Network accessline 78
All endpoints return JSON arrays of objects. Call each endpoint exactly as listed — the path is relative to the resolved base URL. For example, if the base URL is `http://task-env:9009/` and the endpoint is `/api/finance/branches`, - Network accessline 79
listed — the path is relative to the resolved base URL. For example, if the base URL is `http://task-env:9009/` and the endpoint is `/api/finance/branches`, call `http://task-env:9009/api/finance/branches`. - Network accessline 80
base URL is `http://task-env:9009/` and the endpoint is `/api/finance/branches`, - Network accessline 81
call `http://task-env:9009/api/finance/branches`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 19
Use `curl -s "<base_url>/endpoint?params"` for every API call. Parse JSON with `python3 -c` or `python3 -m json.tool`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 51
Fetch all needed endpoints in parallel using curl or similar HTTP calls. All endpoints return JSON arrays or objects.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_02/scripts/validate_template.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_009/fewshot/opencode/fewshot_attempt_03/references/business-rules.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01/SKILL.md
- Network accessline 8
Use this skill when preparing institutional portfolio analytics answers using the shared Asteria Investment Office HTTP environment. The environment is always at `http://task-env:9010/` and requires no credentials. - Network accessline 10
`http://task-env:9010/` and requires no credentials.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01/references/api_endpoints.md
- Network accessline 3
Base URL: `http://task-env:9010/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01/scripts/pearson_corr.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_02/SKILL.md
- Network accessline 21
http://task-env:9010
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 31
A REST API at `http://task-env:9010/` provides the shared book of record. Every
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_03/references/api_surface.md
- Network accessline 3
Base URL: `http://task-env:9010/` (no authentication required)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 15
The Asteria API runs at `http://task-env:9010/` with no authentication.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01/references/api_reference.md
- Network accessline 3
Base URL: `http://task-env:9010/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01/scripts/compute_correlation.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_01/scripts/compute_portfolio_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_02/references/api.md
- Network accessline 3
Base URL: read from the task's `environment_access.md`. The default is `http://task-env:9010/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_02/scripts/quick_validate.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03/SKILL.md
- Network accessline 1
--- name: asteria-portfolio description: Institutional portfolio analytics for the Asteria Investment Office shared environment. Supports energy credit trade construction, international equity correlation review, active allocation view deri - Network accessline 2
name: asteria-portfolio description: Institutional portfolio analytics for the Asteria Investment Office shared environment. Supports energy credit trade construction, international equity correlation review, active allocation view derivati - Network accessline 3
description: Institutional portfolio analytics for the Asteria Investment Office shared environment. Supports energy credit trade construction, international equity correlation review, active allocation view derivation, fixed-income risk re - Network accessline 14
The shared Asteria API (`http://task-env:9010/`) is always the authoritative source. When a local payload date is earlier than the API `as_of_date` (2026-05-29), prefer API values. Report this in any `data_precedence` field as `"current_env
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03/references/api_reference.md
- Network accessline 3
Base URL: `http://task-env:9010/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03/scripts/compute_correlations.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/codex/fewshot_attempt_03/scripts/compute_portfolio_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 1
--- name: asteria-portfolio-risk description: Institutional portfolio risk analysis for the Asteria Investment Office. Covers energy-credit trade strategy, international equity correlation review, active allocation view refresh, fixed-incom - Network accessline 2
name: asteria-portfolio-risk description: Institutional portfolio risk analysis for the Asteria Investment Office. Covers energy-credit trade strategy, international equity correlation review, active allocation view refresh, fixed-income ri - Network accessline 3
description: Institutional portfolio risk analysis for the Asteria Investment Office. Covers energy-credit trade strategy, international equity correlation review, active allocation view refresh, fixed-income risk rebalance, and multi-asset - Network accessline 14
http://task-env:9010/ - Dangerous shell operationline 33
curl -s http://task-env:9010/api/catalog | python3 -m json.tool - Network accessline 33
curl -s http://task-env:9010/api/catalog | python3 -m json.tool
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_01/references/api_reference.md
- Network accessline 3
Base URL: http://task-env:9010/
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_01/scripts/compute_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 22
http://task-env:9010
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/references/api_endpoints.md
- Network accessline 3
Base URL: `http://task-env:9010`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/scripts/compute_correlation.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/scripts/compute_portfolio_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/scripts/resolve_allocation_view.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_03/scripts/correlation.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_03/scripts/weighted_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 12
The shared Asteria Investment Office service is at `http://task-env:9010/`. All endpoints are read-only GETs with no authentication. The service acts as the current book of record for portfolios, instruments, indices, issuers, allocation vi
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_01/api_reference.md
- Network accessline 3
All endpoints at `http://task-env:9010/` are read-only GETs with no authentication.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 8
You are supporting the Asteria Investment Office — a fictional institutional investment manager. Every task follows the same core pattern: read the task prompt, read the local answer-template JSON (which defines the exact output schema), re - Network accessline 16
The Asteria environment runs at `http://task-env:9010/`. No credentials are required. Use `curl` with `-s` to suppress progress output. All endpoints return JSON. - Network accessline 38
Call endpoints with `curl -s http://task-env:9010/<path>`. If the response is large, pipe through Python (`python3 -c "import sys,json; ..."`) to extract the needed fields. Do not assume fields exist; read the response first and then map to
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02/api_reference.md
- Network accessline 3
Base URL: `http://task-env:9010/` - Dangerous shell operationline 110
curl -s http://task-env:9010/api/portfolios/PF-EN-ALTA/holdings | python3 -c " - Network accessline 110
curl -s http://task-env:9010/api/portfolios/PF-EN-ALTA/holdings | python3 -c " - Dangerous shell operationline 120
curl -s http://task-env:9010/api/instruments/bonds | python3 -c " - Network accessline 120
curl -s http://task-env:9010/api/instruments/bonds | python3 -c "
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 22
The API is served at the base URL given in `environment_access.md` (typically `http://task-env:9010/`). No credentials are required. Start every task by reading the API landing page to confirm available endpoints.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_01/references/api-reference.md
- Network accessline 3
All endpoints return JSON. Base URL is given in `environment_access.md` (typically `http://task-env:9010/`). No authentication required.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02/SKILL.md
- Network accessline 26
The base URL is documented in the task's environment context. Read that context (usually `environment_access.md` or a task-group `env/README.md`) for the exact URL and allowed endpoints. A typical base is `http://task-env:9010/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02/references/api.md
- Network accessline 3
This document describes the Asteria Investment Office shared environment service endpoints. The base URL is provided in the task's environment context (typically `http://task-env:9010/`). All endpoints return JSON. - Network accessline 5
context (typically `http://task-env:9010/`). All endpoints return JSON.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02/scripts/correlation.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03/references/api_catalog.md
- Network accessline 3
Base URL: `http://task-env:9010/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03/scripts/pearson_corr.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03/scripts/portfolio_metrics.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 226
peer state from the segment's `peer_states` list. 3. Compute peer medians for each metric. 4. Compare NC to US and NC to peer median on all four dimensions. - Network accessline 227
3. Compute peer medians for each metric. 4. Compare NC to US and NC to peer median on all four dimensions. 5. Determine posture: `continue_approving` when state metrics are stronger - Network accessline 228
4. Compare NC to US and NC to peer median on all four dimensions.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 19
curl -s <TASK_ENV_BASE_URL>/api/<endpoint>
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 22
The base URL arrives as `<TASK_ENV_BASE_URL>`. All endpoints are GET-only. Call them with `curl -s` and pipe through `python3 -m json.tool` for readability. Use `python3 -c` for inline computation.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 62
- Extract NC, US, and peer state rows from the NCUA benchmark table. - Compute peer median for each metric. - Compare NC vs US and NC vs peer median directionally. - Network accessline 63
- Compute peer median for each metric. - Compare NC vs US and NC vs peer median directionally. - Recommend posture based on capacity and external risk signals. - Network accessline 64
- Compare NC vs US and NC vs peer median directionally.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_01/references/api.md
- Network accessline 3
Base URL: supplied by the runner as `<TASK_ENV_BASE_URL>` (defaults to `http://task-env:9011`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_01/references/policy.md
- Network accessline 176
**Peer median**: compute the median of peer state values (exclude NC and US rows from the median set). - Network accessline 182
For NCUA direction: compare NC value to US or peer median; report `"higher"`, `"lower"`, or `"equal"`. - Network accessline 221
1. Fetch segment: `GET /api/credit-union-segments/{segment_id}`. 2. Fetch NCUA benchmarks; extract NC row, US row, and peer state rows. - Network accessline 222
2. Fetch NCUA benchmarks; extract NC row, US row, and peer state rows. - Network accessline 223
3. Compute peer median for each of the four metrics. 4. Compare NC vs US and NC vs peer median (direction only). 5. Posture based on capacity and external risk signals. - Network accessline 224
4. Compare NC vs US and NC vs peer median (direction only).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 132
- `restrained` → `temporarily_pause` 3. **State metrics:** Extract the NCUA row for the segment's `state_code`. Values are integers exactly as reported: `delinquency_bps`, `loan_to_share_pct`, `roaa_bps`, `positive_net_income_pct`. 4. **Pee - Network accessline 133
3. **State metrics:** Extract the NCUA row for the segment's `state_code`. Values are integers exactly as reported: `delinquency_bps`, `loan_to_share_pct`, `roaa_bps`, `positive_net_income_pct`. 4. **Peer comparison:** Use `peer_states` fro - Network accessline 134
4. **Peer comparison:** Use `peer_states` from segment. Compute `nc_vs_us`: compare NC's value against the US row for each of the four metrics → `higher`, `lower`, or `equal`. Compute `nc_vs_peer_median`: take the median of peer states' val - Network accessline 137
7. **Interpretation:** Fill each field from template choices: - `capacity_status`: based on segment's `quarterly_capacity` and internal context. - `external_risk_status`: based on peer comparison results. If NC is mostly worse than pe - Network accessline 138
- `capacity_status`: based on segment's `quarterly_capacity` and internal context. - `external_risk_status`: based on peer comparison results. If NC is mostly worse than peers and US → `weaker_than_national_and_peers`. If mixed → `mixed_ - Network accessline 139
- `external_risk_status`: based on peer comparison results. If NC is mostly worse than peers and US → `weaker_than_national_and_peers`. If mixed → `mixed_vs_national_and_peers`. If mostly better → `stronger_than_national_and_peers`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_03/references/credit_methodology.md
- Network accessline 157
Compare NC vs US direction, and NC vs peer median direction, on four metrics: - Network accessline 173
- capacity_status: compare quarterly_capacity to current_outstanding. If capacity is well above outstanding -> capacity_available. If tight -> capacity_constrained. - external_risk_status: from NCUA comparison. If NC is worse than both US - Network accessline 174
well above outstanding -> capacity_available. If tight -> capacity_constrained. - external_risk_status: from NCUA comparison. If NC is worse than both US and peers on most metrics -> weaker_than_national_and_peers. - Network accessline 175
- external_risk_status: from NCUA comparison. If NC is worse than both US and peers on
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 286
for each of the four metrics, take the median across the peer states. If an even number of peer states, average the middle two values. 5. Compare NC values to US and peer median on `delinquency_bps`, - Network accessline 287
even number of peer states, average the middle two values. 5. Compare NC values to US and peer median on `delinquency_bps`, `loan_to_share_pct`, `roaa_bps`, `positive_net_income_pct`. Directions are - Network accessline 288
5. Compare NC values to US and peer median on `delinquency_bps`,
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 141
For credit union tasks: read the NC state row from `/api/benchmarks/ncua/q1-2025`. Compute the peer median for each metric across the `peer_states` list from the segment endpoint. Compare NC to US and NC to peer median with direction labels
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 2
name: credit-risk-committee description: Bank credit risk committee analysis using a shared credit office REST API. Use this skill whenever the user mentions credit risk committee, branch loan review, rating migration, lending committee all - Network accessline 3
description: Bank credit risk committee analysis using a shared credit office REST API. Use this skill whenever the user mentions credit risk committee, branch loan review, rating migration, lending committee allocation, credit union segmen - Network accessline 4
compatibility: Requires curl and a JSON processor (jq or Python json module). - Network accessline 51
The API base URL is provided as `<TASK_ENV_BASE_URL>` in the prompt. Replace it with the actual URL before making requests. Use `curl -s` and pipe through `jq` or Python for parsing.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_01/references/api-guide.md
- Dangerous shell operationline 18
curl -s <TASK_ENV_BASE_URL>/api/manifest | python3 -c "import sys,json; d=json.load(sys.stdin); print(json.dumps(d,indent=2))" - Network accessline 18
curl -s <TASK_ENV_BASE_URL>/api/manifest | python3 -c "import sys,json; d=json.load(sys.stdin); print(json.dumps(d,indent=2))" - Network accessline 156
Use shell backgrounding or separate curl calls. Example pattern: - Network accessline 159
curl -s "$BASE/api/manifest" > /tmp/manifest.json & - Network accessline 160
curl -s "$BASE/api/policies" > /tmp/policies.json & - Network accessline 161
curl -s "$BASE/api/branches/$BID" > /tmp/branch.json & - Network accessline 162
curl -s "$BASE/api/branches/$BID/metrics" > /tmp/metrics.json & - Network accessline 163
curl -s "$BASE/api/branches/$BID/loans" > /tmp/loans.json & - Network accessline 164
curl -s "$BASE/api/branches/$BID/sector-exposures" > /tmp/sectors.json & - Network accessline 165
curl -s "$BASE/api/benchmarks/fdic/q4-2024" > /tmp/fdic.json &
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_02/SKILL.md
- Network accessline 135
and the US row. Compute peer median: for each metric, take the median of the peer_states' values. 3. Compare NC vs US and NC vs peer_median on all four metrics: delinquency_bps, - Network accessline 136
peer_states' values. 3. Compare NC vs US and NC vs peer_median on all four metrics: delinquency_bps, loan_to_share_pct, roaa_bps, positive_net_income_pct. Direction is higher, - Network accessline 137
3. Compare NC vs US and NC vs peer_median on all four metrics: delinquency_bps, - Network accessline 157
7. **Interpretation**: map capacity_status (capacity_available when quarterly_capacity > 0), external_risk_status (weaker_than_national_and_peers when NC metrics are directionally worse than both US and peer median), - Network accessline 158
quarterly_capacity > 0), external_risk_status (weaker_than_national_and_peers when NC metrics are directionally worse than both US and peer median), risk_tolerance from the segment, and committee_message from the - Network accessline 159
when NC metrics are directionally worse than both US and peer median),
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_02/references/policies.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_02/references/reason_codes.md
- Network accessline 23
| `documentation_gap` | documentation_complete is 0/false and the gap is material. | | `fdic_adverse_variance` | Branch delinquency or NPA ratio substantially exceeds the FDIC benchmark (variance > 10 bps adverse). | | `ncua_peer_weakness` - Network accessline 25
| `ncua_peer_weakness` | NC state metrics are directionally worse than both national and peer median on key dimensions. | - Network accessline 121
Standard triggers for segment posture: ET001, ET002, ET003. Add ET004 when NC delinquency_bps exceeds both US and peer median by a widening margin. - Network accessline 122
NC delinquency_bps exceeds both US and peer median by a widening margin.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 157
Direction comparisons: compare NC value to US value and to peer median. "higher", "lower", or "equal" (use "equal" only when values match exactly).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03/references/task-patterns.md
- Network accessline 129
Get peer_states from segment data. Find rows for each peer state in NCUA data. Compute peer_median for each metric. Compare NC vs US (national row) and NC vs peer_median: "higher", "lower", or "equal".
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03/scripts/compute.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_01/SKILL.md
- Dangerous shell operationline 17
curl -s <TASK_ENV_BASE_URL>/api/manifest | python3 -m json.tool - Network accessline 17
curl -s <TASK_ENV_BASE_URL>/api/manifest | python3 -m json.tool - Dangerous shell operationline 18
curl -s <TASK_ENV_BASE_URL>/api/summary | python3 -m json.tool - Network accessline 18
curl -s <TASK_ENV_BASE_URL>/api/summary | python3 -m json.tool - Dangerous shell operationline 19
curl -s <TASK_ENV_BASE_URL>/api/employees | python3 -m json.tool - Network accessline 19
curl -s <TASK_ENV_BASE_URL>/api/employees | python3 -m json.tool - Dangerous shell operationline 20
curl -s <TASK_ENV_BASE_URL>/api/cases | python3 -m json.tool - Network accessline 20
curl -s <TASK_ENV_BASE_URL>/api/cases | python3 -m json.tool - Dangerous shell operationline 21
curl -s <TASK_ENV_BASE_URL>/api/payroll-ledgers | python3 -m json.tool - Network accessline 21
curl -s <TASK_ENV_BASE_URL>/api/payroll-ledgers | python3 -m json.tool - Dangerous shell operationline 22
curl -s <TASK_ENV_BASE_URL>/api/policies | python3 -m json.tool - Network accessline 22
curl -s <TASK_ENV_BASE_URL>/api/policies | python3 -m json.tool - Dangerous shell operationline 23
curl -s <TASK_ENV_BASE_URL>/api/audit | python3 -m json.tool - Network accessline 23
curl -s <TASK_ENV_BASE_URL>/api/audit | python3 -m json.tool - Dangerous shell operationline 24
curl -s <TASK_ENV_BASE_URL>/api/recruitment | python3 -m json.tool - Network accessline 24
curl -s <TASK_ENV_BASE_URL>/api/recruitment | python3 -m json.tool - Dangerous shell operationline 25
curl -s <TASK_ENV_BASE_URL>/api/documents | python3 -m json.tool - Network accessline 25
curl -s <TASK_ENV_BASE_URL>/api/documents | python3 -m json.tool - Dangerous shell operationline 26
curl -s <TASK_ENV_BASE_URL>/api/messages | python3 -m json.tool - Network accessline 26
curl -s <TASK_ENV_BASE_URL>/api/messages | python3 -m json.tool - Dangerous shell operationline 27
curl -s <TASK_ENV_BASE_URL>/api/notifications | python3 -m json.tool - Network accessline 27
curl -s <TASK_ENV_BASE_URL>/api/notifications | python3 -m json.tool - Dangerous shell operationline 33
curl -s <TASK_ENV_BASE_URL>/api/cases/{case_id} | python3 -m json.tool - Network accessline 33
curl -s <TASK_ENV_BASE_URL>/api/cases/{case_id} | python3 -m json.tool - Dangerous shell operationline 34
curl -s <TASK_ENV_BASE_URL>/api/policies/{policy_id} | python3 -m json.tool - Network accessline 34
curl -s <TASK_ENV_BASE_URL>/api/policies/{policy_id} | python3 -m json.tool - Dangerous shell operationline 35
curl -s <TASK_ENV_BASE_URL>/api/audit/{audit_id} | python3 -m json.tool - Network accessline 35
curl -s <TASK_ENV_BASE_URL>/api/audit/{audit_id} | python3 -m json.tool
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_03/references/api_reference.md
- Network accessline 3
Base URL: `<TASK_ENV_BASE_URL>` (from prompt, typically `http://task-env:9012/`)
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_01/references/api-catalog.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 94
The console base URL is `<TASK_ENV_BASE_URL>`. All endpoints return JSON arrays or objects. No authentication headers needed for API calls. Use `curl -s`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_03/references/api_endpoints.md
- Network accessline 5
All endpoints return JSON. No authentication headers needed for API calls. Use `curl -s` to fetch data.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 82
def fetch(url): - Network accessline 83
with urllib.request.urlopen(url) as r: - Network accessline 88
# Fetch all collections employees = fetch(f"{base}/api/employees") - Network accessline 89
employees = fetch(f"{base}/api/employees") - Network accessline 90
cases = fetch(f"{base}/api/cases") - Network accessline 91
ledgers = fetch(f"{base}/api/payroll-ledgers") - Network accessline 92
policies = fetch(f"{base}/api/policies") - Network accessline 93
docs = fetch(f"{base}/api/documents") - Network accessline 94
msgs = fetch(f"{base}/api/messages") - Network accessline 95
audits = fetch(f"{base}/api/audit") - Network accessline 96
recruitment = fetch(f"{base}/api/recruitment") - Network accessline 97
notifications = fetch(f"{base}/api/notifications") - Network accessline 99
# Filter by employee_id or case_id emp_ledgers = [l for l in ledgers if l["employee_id"] == target_emp_id] case_detail = fetch(f"{base}/api/cases/{target_case_id}") - Network accessline 100
emp_ledgers = [l for l in ledgers if l["employee_id"] == target_emp_id] case_detail = fetch(f"{base}/api/cases/{target_case_id}") emp_audits = [a for a in audits if a["employee_id"] == target_emp_id or a["case_id"] == target_case_id] - Network accessline 101
case_detail = fetch(f"{base}/api/cases/{target_case_id}")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 25
All endpoints are rooted at the configured runner URL, typically `<TASK_ENV_BASE_URL>` or `http://task-env:9012/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03/references/api-details.md
- Network accessline 296
curl -s $BASE/api/payroll-ledgers | jq '[.[] | select(.employee_id=="EMP-104" and .record_type=="Leave assignment")]' - Network accessline 301
curl -s $BASE/api/payroll-ledgers | jq '[.[] | select(.employee_id=="EMP-104" and .record_type=="Salary assignment")]' - Network accessline 306
curl -s $BASE/api/payroll-ledgers | jq '[.[] | select(.employee_id=="EMP-104" and .record_type=="Leave assignment" and .status != "Draft" and .status != "Superseded")]' - Network accessline 314
curl -s $BASE/api/audit | jq '[.[] | select(.case_id=="CASE-445" or .employee_id=="EMP-255")]'
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 17
# List all patients curl -s "<TASK_ENV_BASE_URL>/patients" - Network accessline 18
curl -s "<TASK_ENV_BASE_URL>/patients" - Network accessline 20
# Get one patient (composite: demographics, coverage, PBM, pharmacy, lifestyle, documents, referrals, transfers, rosters, program candidates) curl -s "<TASK_ENV_BASE_URL>/patients/<patient_id>" - Network accessline 21
curl -s "<TASK_ENV_BASE_URL>/patients/<patient_id>" - Network accessline 23
# Get patient chart (chart_artifacts, active_problems, meds_allergies, clinical_history, recent_vitals_labs) curl -s "<TASK_ENV_BASE_URL>/chart/<patient_id>" - Network accessline 24
curl -s "<TASK_ENV_BASE_URL>/chart/<patient_id>" - Network accessline 26
# List referrals for a batch curl -s "<TASK_ENV_BASE_URL>/referrals?batch_id=<batch_id>" - Network accessline 27
curl -s "<TASK_ENV_BASE_URL>/referrals?batch_id=<batch_id>" - Network accessline 29
# Get a single referral curl -s "<TASK_ENV_BASE_URL>/referrals/<referral_id>" - Network accessline 30
curl -s "<TASK_ENV_BASE_URL>/referrals/<referral_id>" - Network accessline 32
# List transfers for a batch curl -s "<TASK_ENV_BASE_URL>/transfers?batch_id=<batch_id>" - Network accessline 33
curl -s "<TASK_ENV_BASE_URL>/transfers?batch_id=<batch_id>" - Network accessline 35
# List all documents (or filter with query params) curl -s "<TASK_ENV_BASE_URL>/documents" - Network accessline 36
curl -s "<TASK_ENV_BASE_URL>/documents" - Network accessline 38
# Get ICD metadata curl -s "<TASK_ENV_BASE_URL>/icd/<code>" - Network accessline 39
curl -s "<TASK_ENV_BASE_URL>/icd/<code>" - Network accessline 41
# List pharmacies curl -s "<TASK_ENV_BASE_URL>/pharmacies" - Network accessline 42
curl -s "<TASK_ENV_BASE_URL>/pharmacies" - Network accessline 44
# Get program candidates curl -s "<TASK_ENV_BASE_URL>/programs/<program_code>/candidates" - Network accessline 45
curl -s "<TASK_ENV_BASE_URL>/programs/<program_code>/candidates" - Network accessline 47
# Run a read-only SQL query curl -s -X POST "<TASK_ENV_BASE_URL>/query" \ - Network accessline 48
curl -s -X POST "<TASK_ENV_BASE_URL>/query" \ - Network accessline 78
Always quote string literals carefully in curl (escape single quotes with `'\''`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 47
controlled vocabularies, ordering rules, and required keys. 3. Replace `<TASK_ENV_BASE_URL>` with the actual base URL given in the task environment. Use `curl -s` for every GET call. - Network accessline 48
3. Replace `<TASK_ENV_BASE_URL>` with the actual base URL given in the task environment. Use `curl -s` for every GET call. 4. Collect all referenced records: fetch the collection endpoint first, then - Network accessline 49
environment. Use `curl -s` for every GET call.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01/references/api_reference.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01/scripts/query_portal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 48
def get(base_url: str, endpoint: str) -> dict | list: url = f"{base_url.rstrip('/')}/{endpoint.lstrip('/')}" with urllib.request.urlopen(url) as resp: - Network accessline 49
url = f"{base_url.rstrip('/')}/{endpoint.lstrip('/')}" with urllib.request.urlopen(url) as resp: return json.loads(resp.read()) - Network accessline 50
with urllib.request.urlopen(url) as resp: - Network accessline 55
url = f"{base_url.rstrip('/')}/query" body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request(url, data=body, headers={"Content-Type": "application/json"}, method="POST") - Network accessline 56
body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request(url, data=body, headers={"Content-Type": "application/json"}, method="POST") with urllib.request.urlopen(req) as resp: - Network accessline 57
req = urllib.request.Request(url, data=body, headers={"Content-Type": "application/json"}, method="POST") - Network accessline 58
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_03/scripts/validate_output.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_01/references/environment.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 33
curl -s -X POST <BASE_URL>/sql/query \
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_03/scripts/query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 24
"--base-url", required=True, help="Task environment base URL, e.g. http://task-env:9014", - Network accessline 25
required=True, help="Task environment base URL, e.g. http://task-env:9014", ) - Network accessline 26
help="Task environment base URL, e.g. http://task-env:9014", - Network accessline 38
req = urllib.request.Request( - Network accessline 48
try: with urllib.request.urlopen(req) as resp: - Network accessline 49
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 15
# Schema discovery (always run first) curl -s <BASE_URL>/api/tables | python3 -m json.tool - Dangerous shell operationline 16
curl -s <BASE_URL>/api/tables | python3 -m json.tool - Network accessline 16
curl -s <BASE_URL>/api/tables | python3 -m json.tool - Network accessline 18
# SQL queries (primary data access) curl -s -X POST <BASE_URL>/sql/query \ - Network accessline 19
curl -s -X POST <BASE_URL>/sql/query \ - Network accessline 212
Use curl with the POST endpoint. The JSON body must have the key `sql`: - Network accessline 215
curl -s -X POST <BASE_URL>/sql/query \ - Network accessline 223
When joining across tables, use standard SQL JOIN syntax. The database supports basic SQL: SELECT, FROM, JOIN, WHERE, ORDER BY, GROUP BY. String literals use single quotes. Escape single quotes in shell by using the `'\''` pattern within si
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_03/references/sql-helper.md
- Network accessline 3
Reusable pattern for querying the Northstar payer-operations SQL endpoint. Use this when curl escaping becomes unwieldy for multi-condition queries. - Network accessline 5
## Curl Quoting Reference - Network accessline 11
curl -s -X POST <BASE_URL>/sql/query \ - Network accessline 17
The `'\''` sequence inside single-quoted curl `-d` strings closes the outer single-quote, inserts a literal single quote, then reopens. For queries with many string conditions, prefer the Python approach below. - Network accessline 27
def query(sql): r = requests.post(f"{BASE}/sql/query", headers=HEADERS, json={"sql": sql}) - Network accessline 28
r = requests.post(f"{BASE}/sql/query", headers=HEADERS, json={"sql": sql})
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_02/references/api_endpoints.md
- Network accessline 70
curl -s "$BASE/api/patients/P-XXXXX" & - Network accessline 71
curl -s "$BASE/api/patients/P-XXXXX/conditions" & - Network accessline 72
curl -s "$BASE/api/patients/P-XXXXX/medications" & - Network accessline 73
curl -s "$BASE/api/patients/P-XXXXX/allergies" & - Network accessline 74
curl -s "$BASE/api/patients/P-XXXXX/encounters" & - Network accessline 75
curl -s "$BASE/api/patients/P-XXXXX/documents" & - Network accessline 76
curl -s "$BASE/api/patients/P-XXXXX/immunizations" & - Network accessline 77
curl -s "$BASE/api/patients/P-XXXXX/disclosures" & - Network accessline 84
# Check existence (404 = invalid) curl -s -o /dev/null -w "%{http_code}" "$BASE/api/icd10/I50.32" - Network accessline 85
curl -s -o /dev/null -w "%{http_code}" "$BASE/api/icd10/I50.32" - Network accessline 87
# Get details curl -s "$BASE/api/icd10/I50.32" - Network accessline 88
curl -s "$BASE/api/icd10/I50.32" - Network accessline 95
curl -s "$BASE/api/providers/PRV-CARD-020" - Network accessline 112
# Get all referrals curl -s "$BASE/api/referrals" | jq '[.[] | select(.batch_id == "MAR26-ORTHO-A")]' - Network accessline 113
curl -s "$BASE/api/referrals" | jq '[.[] | select(.batch_id == "MAR26-ORTHO-A")]'
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 14
1. Read the prompt and `answer_template.json`. 2. Resolve `<TASK_ENV_BASE_URL>` to `http://task-env:9015`. - Network accessline 15
2. Resolve `<TASK_ENV_BASE_URL>` to `http://task-env:9015`. - Network accessline 33
- Use `curl -s` to call endpoints. Always pipe through `python3 -m json.tool` or `jq` for readable parsing.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03/references/api-endpoints.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03/references/workflows.md
- Dangerous shell operationline 202
curl -s http://task-env:9015/api/patients/P-NNNNN | python3 -m json.tool > /tmp/patient_a.json & - Network accessline 202
curl -s http://task-env:9015/api/patients/P-NNNNN | python3 -m json.tool > /tmp/patient_a.json & - Dangerous shell operationline 203
curl -s http://task-env:9015/api/patients/P-NNNNN | python3 -m json.tool > /tmp/patient_b.json & - Network accessline 203
curl -s http://task-env:9015/api/patients/P-NNNNN | python3 -m json.tool > /tmp/patient_b.json & - Network accessline 210
# Get active condition keys for a patient, sorted curl -s http://task-env:9015/api/patients/P-NNNNN/conditions | jq -r '[.[] | select(.status=="active") | .normalized_key] | sort | .[]' - Network accessline 211
[Credential-related source omitted] - Access to secretsline 211
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 14
Use `curl -s --connect-timeout 10 --max-time 30` for every API call. Prefer the helper script at [scripts/fetch.sh](scripts/fetch.sh) which wraps curl with retry logic and error handling; invoke it as `bash /work/skill/scripts/fetch.sh <end - Network accessline 116
A reusable curl wrapper for the EHR API. Usage: - Network accessline 124
Read [scripts/fetch.sh](scripts/fetch.sh) for the full implementation. Use it for every individual API call to avoid repetitive curl boilerplate.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_03/scripts/fetch.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 4
# # Reads TASK_ENV_BASE_URL from the environment. Set it before calling: # export TASK_ENV_BASE_URL="http://task-env:9015" - Network accessline 5
# Reads TASK_ENV_BASE_URL from the environment. Set it before calling: # export TASK_ENV_BASE_URL="http://task-env:9015" # - Network accessline 6
# export TASK_ENV_BASE_URL="http://task-env:9015" - Network accessline 33
HTTP_CODE=$(curl -s -w '%{http_code}' -o /tmp/fetch_response.json \ - Network accessline 37
if [ "$HTTP_CODE" = "000" ] || [ "${HTTP_CODE:0:1}" = "5" ]; then HTTP_CODE=$(curl -s -w '%{http_code}' -o /tmp/fetch_response.json \ - Network accessline 38
HTTP_CODE=$(curl -s -w '%{http_code}' -o /tmp/fetch_response.json \
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 29
Read the environment access file (`environment_access.md` or similarly named) to learn the base URL and the allowed business endpoints. The base URL appears as `http://task-env:<PORT>/` with endpoints like `/api/patients`, `/api/cases/{case
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 20
- **case_id**: The target case identifier (e.g. CASE-RESP-102). - **TASK_ENV_BASE_URL**: The clinic runtime base URL. When the prompt says <TASK_ENV_BASE_URL>, substitute http://task-env:9016/ unless the environment access documentation pro - Network accessline 21
- **TASK_ENV_BASE_URL**: The clinic runtime base URL. When the prompt says <TASK_ENV_BASE_URL>, substitute http://task-env:9016/ unless the environment access documentation provides a different value.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_02/references/api_endpoints.md
- Network accessline 3
Base URL: `http://task-env:9016/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_02/scripts/fetch_case.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
url = f"{base_url.rstrip('/')}/api/cases/{case_id}" try: with urllib.request.urlopen(url, timeout=10) as resp: - Network accessline 19
try: with urllib.request.urlopen(url, timeout=10) as resp: return json.loads(resp.read()) - Network accessline 20
with urllib.request.urlopen(url, timeout=10) as resp: - Network accessline 35
parser.add_argument( "--base-url", default="http://task-env:9016", - Network accessline 36
"--base-url", default="http://task-env:9016", help="Clinic runtime base URL (default: http://task-env:9016)", - Network accessline 37
default="http://task-env:9016", - Network accessline 38
help="Clinic runtime base URL (default: http://task-env:9016)",
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_03/scripts/fetch_case.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 19
try: with urllib.request.urlopen(req, timeout=15) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 20
with urllib.request.urlopen(req, timeout=15) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_03/scripts/probe_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 20
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 21
with urllib.request.urlopen(req) as resp: - Network accessline 27
data = json.dumps(body).encode() req = urllib.request.Request( - Network accessline 28
req = urllib.request.Request( - Network accessline 29
url, data=data, headers={"Content-Type": "application/json", "Accept": "application/json"} ) with urllib.request.urlopen(req) as resp: - Network accessline 30
) with urllib.request.urlopen(req) as resp: return json.loads(resp.read().decode()) - Network accessline 31
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_01/scripts/query_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 9
[Credential-related source omitted] - Network accessline 19
[Credential-related source omitted] - Access to secretsline 19
[Credential-related source omitted] - Network accessline 20
url = f"{base_url}/api/query" data = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( - Network accessline 21
data = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( url, - Network accessline 22
req = urllib.request.Request( - Access to secretsline 27
[Credential-related source omitted] - Network accessline 30
) try: with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 31
try: with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 32
with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 55
parser.add_argument( "--base-url", default="http://task-env:9017", - Network accessline 56
"--base-url", default="http://task-env:9017", help="Hub base URL (default: http://task-env:9017)", - Network accessline 57
default="http://task-env:9017", - Network accessline 58
help="Hub base URL (default: http://task-env:9017)", - Access to secretsline 80
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_02/references/sql_query_patterns.md
- Network accessline 8
curl "$BASE/api/schema"
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02/references/hub_api_reference.md
- Access to secretsline 7
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 181
curl -s -X POST "$BASE_URL/api/query" \
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 26
The task prompt and payloads include a base URL (often referenced as `<TASK_ENV_BASE_URL>` or a literal `http://` URL). Use the `GET /` and `GET /api/schema` endpoints to confirm the available endpoints and learn the exact field names used
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_01/scripts/payment_calc.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 26
- Base URL: use the value from the prompt verbatim (typically `http://task-env:9019`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_03/SKILL.md
- Network accessline 17
curl -s <TASK_ENV_BASE_URL>/api/policies - Network accessline 25
[Credential-related source omitted] - Dangerous shell operationline 200
- **[correspondence_filter.py](scripts/correspondence_filter.py)** — Reads a JSON array of correspondence records from stdin and outputs a sorted, deduplicated list of IDs whose status indicates stale, unverified, unconfirmed, or pending. U - Network accessline 200
- **[correspondence_filter.py](scripts/correspondence_filter.py)** — Reads a JSON array of correspondence records from stdin and outputs a sorted, deduplicated list of IDs whose status indicates stale, unverified, unconfirmed, or pending. U
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_03/scripts/correspondence_filter.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 9
Usage: cat correspondence.json | python3 correspondence_filter.py curl -s $BASE/api/contractor/correspondence | python3 correspondence_filter.py - Network accessline 10
cat correspondence.json | python3 correspondence_filter.py curl -s $BASE/api/contractor/correspondence | python3 correspondence_filter.py """ - Dangerous shell operationline 11
curl -s $BASE/api/contractor/correspondence | python3 correspondence_filter.py - Network accessline 11
curl -s $BASE/api/contractor/correspondence | python3 correspondence_filter.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_02/scripts/fetch_data.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 63
url = base_url.rstrip("/") + path req = urllib.request.Request(url, method="GET") - Network accessline 64
req = urllib.request.Request(url, method="GET") - Network accessline 66
req.add_header("X-Task-Token", token) try: with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 67
try: with urllib.request.urlopen(req, timeout=30) as resp: body = resp.read().decode("utf-8") - Network accessline 68
with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 83
def main() -> None: parser = argparse.ArgumentParser(description="Fetch licensing data from the task environment.") parser.add_argument("--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9019"), - Network accessline 84
parser = argparse.ArgumentParser(description="Fetch licensing data from the task environment.") parser.add_argument("--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9019"), help="Base UR - Network accessline 85
[Credential-related source omitted] - Access to secretsline 85
[Credential-related source omitted] - Network accessline 86
help="Base URL of the licensing environment (default: http://task-env:9019 or $TASK_ENV_BASE_URL)") - Access to secretsline 87
[Credential-related source omitted] - Network accessline 88
[Credential-related source omitted] - Access to secretsline 88
[Credential-related source omitted] - Network accessline 89
parser.add_argument("--domain", choices=["all", "contractor", "liquor", "alcohol", "renewal"], default="all", help="Which domain endpoints to fetch (default: all)") args = parser.parse_args() - Network accessline 90
help="Which domain endpoints to fetch (default: all)")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 22
The task environment base URL is `http://task-env:9019`. Authentication for `POST /api/sql` uses header `X-Task-Token` with the value from the prompt or environment instructions. No other endpoints require authentication.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_02/references/api_patterns.md
- Network accessline 18
# Example: parallel fetch for contractor review curl -s "${BASE}/api/policies" & - Network accessline 19
curl -s "${BASE}/api/policies" & - Network accessline 20
curl -s "${BASE}/api/contractor/applications" & - Network accessline 21
curl -s "${BASE}/api/contractor/bonds" & - Network accessline 22
curl -s "${BASE}/api/contractor/insurance" & - Network accessline 23
curl -s "${BASE}/api/contractor/license-history" & - Network accessline 24
curl -s "${BASE}/api/contractor/violations" & - Network accessline 25
curl -s "${BASE}/api/contractor/correspondence" & - Network accessline 26
curl -s "${BASE}/api/contractor/inspections" & - Network accessline 35
curl -s -X POST "${BASE}/api/sql" \
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_02/SKILL.md
- Network accessline 2
name: deal-workbench description: Use the M&A deal workbench REST API to review acquisition agreements, produce structured legal deliverables (issue registers, closing-package analyses, committee escalation memos, transition reviews, deviat - Network accessline 3
description: Use the M&A deal workbench REST API to review acquisition agreements, produce structured legal deliverables (issue registers, closing-package analyses, committee escalation memos, transition reviews, deviation matrices), and cr - Network accessline 4
compatibility: This skill assumes a running M&A deal workbench server. The base URL can be provided explicitly or through the `<TASK_ENV_BASE_URL>` sentinel. No local tooling is required beyond an HTTP client (curl or equivalent). - Network accessline 41
2. **Pull the full deal data.** Fetch all relevant endpoints **in parallel** when requests are independent (for example, `/deals/<id>`, `/deals/<id>/terms`, `/playbooks/<id>/rules` can all be fetched simultaneously). Use curl with `--parall
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_03/references/workbench_api.md
- Network accessline 8
All GET endpoints return JSON. Use `curl -s` or equivalent without
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_020/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 12
All endpoints live under the base URL provided in the task prompt (e.g. `http://task-env:9020`). Call `GET` for reads and `POST /api/query` with the read-only token for cross-table SQL.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_02/references/api_endpoints.md
- Network accessline 8
base_url: http://task-env:9021/
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_02/scripts/query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 9
Or import and use programmatically: from query import load_all rows = load_all("http://task-env:9021", "transactions", "fuel_purchases_2026_01") - Network accessline 10
from query import load_all rows = load_all("http://task-env:9021", "transactions", "fuel_purchases_2026_01") """ - Network accessline 11
rows = load_all("http://task-env:9021", "transactions", "fuel_purchases_2026_01") - Network accessline 25
def _get_json(url: str) -> dict: req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 26
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 27
with urllib.request.urlopen(req) as resp: - Network accessline 31
def _post_json(url: str, body: dict) -> dict: data = json.dumps(body).encode("utf-8") req = urllib.request.Request( - Network accessline 32
data = json.dumps(body).encode("utf-8") req = urllib.request.Request( url, - Network accessline 33
req = urllib.request.Request( - Network accessline 37
method="POST", ) with urllib.request.urlopen(req) as resp: - Network accessline 38
) with urllib.request.urlopen(req) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(req) as resp: - Network accessline 53
Args: base_url: The hub base URL (e.g., http://task-env:9021). - Network accessline 54
base_url: The hub base URL (e.g., http://task-env:9021).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_03/scripts/query_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 28
req = urllib.request.Request( - Network accessline 37
try: with urllib.request.urlopen(req, timeout=60) as resp: - Network accessline 38
with urllib.request.urlopen(req, timeout=60) as resp: - Network accessline 63
parser = argparse.ArgumentParser(description="Paginated query helper for Fleet DQ Hub") parser.add_argument("base_url", help="Hub base URL (e.g., http://task-env:9021)") - Network accessline 64
parser.add_argument("base_url", help="Hub base URL (e.g., http://task-env:9021)")
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_01/scripts/README.md
- Network accessline 33
# Fetch all fuel transactions python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/transactions/fuel" > /tmp/fuel.json - Network accessline 34
python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/transactions/fuel" > /tmp/fuel.json - Network accessline 36
# Fetch all freight charges python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/transactions/freight" > /tmp/freight.json - Network accessline 37
python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/transactions/freight" > /tmp/freight.json - Network accessline 39
# Fetch all maintenance events python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/maintenance/events" > /tmp/events.json - Network accessline 40
python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/maintenance/events" > /tmp/events.json - Network accessline 42
# Query with filtering echo '{"collection_id":"some_collection","filters":[],"limit":500}' > /tmp/query.json python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/query" --query /tmp/query.json > /tmp/results.json - Network accessline 44
python3 /work/skill/scripts/fetch_all.py "http://task-env:9021/api/query" --query /tmp/query.json > /tmp/results.json
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_01/scripts/fetch_all.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
if post_body is not None: req_data = json.dumps(post_body).encode("utf-8") req = urllib.request.Request(req_url, data=req_data, method="POST") - Network accessline 19
req_data = json.dumps(post_body).encode("utf-8") req = urllib.request.Request(req_url, data=req_data, method="POST") req.add_header("Content-Type", "application/json") - Network accessline 20
req = urllib.request.Request(req_url, data=req_data, method="POST") - Network accessline 21
req.add_header("Content-Type", "application/json") else: req = urllib.request.Request(req_url, method="GET") - Network accessline 23
req = urllib.request.Request(req_url, method="GET") - Network accessline 25
try: with urllib.request.urlopen(req) as resp: - Network accessline 26
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/check_contamination.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/verify_links.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/write_refs.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_02/SKILL.md
- Network accessline 14
- **Base URL**: the value of `<TASK_ENV_BASE_URL>` from the task prompt, defaulting to `http://task-env:9022/`
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02/scripts/query_atlas.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 45
### curl Patterns - Network accessline 52
# Health curl -s -H "$AUTH" -H "$CT" "$BASE/health" - Network accessline 53
curl -s -H "$AUTH" -H "$CT" "$BASE/health" - Network accessline 55
# Schema (save for reference) curl -s -H "$AUTH" -H "$CT" "$BASE/api/schema" > /tmp/schema.json - Network accessline 56
curl -s -H "$AUTH" -H "$CT" "$BASE/api/schema" > /tmp/schema.json - Network accessline 58
# Data dictionary curl -s -H "$AUTH" -H "$CT" "$BASE/api/data-dictionary" > /tmp/dict.json - Network accessline 59
curl -s -H "$AUTH" -H "$CT" "$BASE/api/data-dictionary" > /tmp/dict.json - Network accessline 61
# Read-only SQL curl -s -H "$AUTH" -H "$CT" -d '{"sql": "SELECT ..."}' "$BASE/api/sql" - Network accessline 62
curl -s -H "$AUTH" -H "$CT" -d '{"sql": "SELECT ..."}' "$BASE/api/sql" - Network accessline 64
# Correction audit curl -s -H "$AUTH" -H "$CT" "$BASE/api/correction-audit" - Network accessline 65
curl -s -H "$AUTH" -H "$CT" "$BASE/api/correction-audit"
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 10
Every task provides a base URL in a placeholder like `<TASK_ENV_BASE_URL>` or a literal URL such as `http://task-env:9022/`. Use that value throughout. The authentication is already wired:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 12
The API base URL is provided in the prompt as `<TASK_ENV_BASE_URL>`. Common staging value is `http://task-env:9022`. - Dangerous shell operationline 59
[Credential-related source omitted] - Network accessline 59
[Credential-related source omitted] - Dangerous shell operationline 60
[Credential-related source omitted] - Network accessline 60
[Credential-related source omitted] - Network accessline 76
[Credential-related source omitted] - Network accessline 77
[Credential-related source omitted] - Network accessline 78
[Credential-related source omitted] - Network accessline 133
- [references/api-patterns.md](references/api-patterns.md) — curl patterns, common SQL snippets, and transaction format
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01/references/api-patterns.md
- Network accessline 6
BASE="${TASK_ENV_BASE_URL:-http://task-env:9022}" - Network accessline 13
# Schema: tables, columns, types, foreign keys curl -s -H "$AUTH" "$BASE/api/schema" > /tmp/schema.json - Network accessline 14
curl -s -H "$AUTH" "$BASE/api/schema" > /tmp/schema.json - Network accessline 16
# Data dictionary: business meanings for columns and enum values curl -s -H "$AUTH" "$BASE/api/data-dictionary" > /tmp/dictionary.json - Network accessline 17
curl -s -H "$AUTH" "$BASE/api/data-dictionary" > /tmp/dictionary.json - Network accessline 38
curl -s -X POST \ - Network accessline 107
curl -s -X POST \ - Dangerous shell operationline 131
curl -s -H "$AUTH" "$BASE/api/correction-audit" | python3 -m json.tool - Network accessline 131
curl -s -H "$AUTH" "$BASE/api/correction-audit" | python3 -m json.tool - Network accessline 144
def query(sql): result = subprocess.run([ "curl", "-s", "-X", "POST", - Network accessline 145
[Credential-related source omitted] - Dangerous shell operationline 170
if eval(tier["condition"]):
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01/scripts/compute.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 17
BASE = "http://task-env:9022" - Network accessline 23
result = subprocess.run( ["curl", "-s", "-H", AUTH_HEADER, f"{BASE}{path}"], - Network accessline 33
result = subprocess.run( [ "curl", "-s", "-X", "POST", - Network accessline 50
result = subprocess.run( [ "curl", "-s", "-X", "POST", - Dangerous shell operationline 106
if eval(condition, {"__builtins__": {}}, context):
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02/references/schemas.md
- Network accessline 297
curl -s -X POST "<BASE>/api/sql/transaction" \
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01/references/portal.md
- Network accessline 7
Provided by the task environment, typically `http://task-env:9023/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01/references/protocols.md
- Dangerous shell operationline 149
From unrounded baseline a, b, SE_b, and residual df, compute magnitude = SE_b * sqrt(df * rY * rM / (1 - rM)). For each declared direction use adjusted_b = b - s*magnitude, adjusted_indirect = a * adjusted_b, adjusted_direct = total - adjus
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01/scripts/download.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_03/references/api.md
- Network accessline 3
Base URL: http://task-env:9023/
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_03/references/methods.md
- Dangerous shell operationline 506
- magnitude = SE_b * sqrt(df * rY * rM / (1 - rM))
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_01/references/statistical_modules.md
- Dangerous shell operationline 775
magnitude = SE_b * sqrt(df * rY * rM / (1 - rM))
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02/SKILL.md
- Dangerous shell operationline 77
7. **partial_r2_sensitivity** — From unrounded baseline `a`, `b`, `SE_b`, and residual `df`, compute `magnitude = SE_b * sqrt(df * rY * rM / (1-rM))`. For each declared direction, `adjusted_b = b - s*magnitude`, `adjusted_indirect = a*adjus
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02/references/statistical_methods.md
- Dangerous shell operationline 321
magnitude = SE_b * sqrt(df * rY * rM / (1 - rM)) - Dangerous shell operationline 322
where rY = outcome confounder partial R^2, rM = mediator confounder partial R^2. - Dangerous shell operationline 330
Tipping point: solve for r^2 where b - SE_b * sqrt(df * r^2 * r^2 / (1 - r^2)) = 0 (equal-strength case rY = rM = r^2).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/references/portal_api.md
- Network accessline 5
`{TASK_ENV_BASE_URL}` — provided in the task prompt (e.g. `http://task-env:9023/`).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/cluster.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/conformal.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/ols.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/pca.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/prng.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/ridge.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_01/references/portal_api.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_01/scripts/download_portal_data.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_02/references/api-endpoints.md
- Network accessline 5
The prompt provides `<TASK_ENV_BASE_URL>`. This is typically `http://task-env:9023/`.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 24
The portal lives at the base URL declared in the task environment (typically `http://task-env:9023/`). See [references/portal-api.md](references/portal-api.md) for the complete endpoint reference, CSV schemas, and download patterns. - Network accessline 28
curl -s "BASE_URL/download?dataset=<dataset>&format=csv&<filters>"
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_03/references/portal-api.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_01/SKILL.md
- Dangerous shell operationline 159
From baseline path-a coefficient a, path-b coefficient b, SE_b, and residual df: magnitude = SE_b * sqrt(df * rY * rM / (1 - rM)). For each direction: adjusted_b = b - sign * magnitude, adjusted_indirect = a * adjusted_b, adjusted_direct =
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_01/references/sensitivity_surface.md
- Dangerous shell operationline 14
For declared R2 values rM (confounder-mediator) and rY (confounder-outcome): - Dangerous shell operationline 15
- magnitude = SE_b * sqrt(df * rY * rM / (1 - rM)) - Dangerous shell operationline 33
The equal-strength tipping R2 is the value of r2 (= rM = rY) such that the POSITIVE adjusted_indirect crosses zero (changes sign from the baseline): - Dangerous shell operationline 42
- Check all rows with direction = POSITIVE, rM <= 0.08, rY <= 0.08.
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 15
Read `environment_access.md` in the workspace root. It provides the base URL (e.g. `http://task-env:9023/`) and the list of allowed endpoints. Use GET requests only; - Network accessline 16
`http://task-env:9023/`) and the list of allowed endpoints. Use GET requests only; - Dangerous shell operationline 222
coefficient b, SE_b, and residual df: magnitude = SE_b * sqrt(df * rY * rM / (1 -
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_02/protocols/county_mediation_v1.md
- Dangerous shell operationline 123
sqrt(df * rY * rM / (1 - rM)).
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 26
Base URL from the task prompt (typically `http://task-env:9023/`). Endpoints: - Dangerous shell operationline 638
magnitude = SE_b * sqrt(df * rY * rM / (1 - rM)) - Dangerous shell operationline 652
Enumerate the complete surface in declared R2 row-major order (rM ascending, - Dangerous shell operationline 655
**Tipping point**: The R2 value (assuming equal strength rM = rY = r2) at
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/adjusted_rand_index.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/conformal.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/coordinate_descent.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/jacobi_eigen.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/pcg32.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/pseudoinverse_helpers.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/shapley.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_03/xorshift32.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_01/references/modules.md
- Dangerous shell operationline 398
magnitude = SE_b * sqrt(df * rY * rM / (1 - rM))
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_02/scripts/compute.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_03/references/methods.md
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_03/scripts/query_portfolio.py
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 24
TASK_ENV_BASE_URL=http://task-env:9024 python scripts/query_env.py work-items
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_01/scripts/query_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 4
Reads TASK_ENV_BASE_URL from the environment (default: http://task-env:9024). - Network accessline 26
[Credential-related source omitted] - Access to secretsline 26
[Credential-related source omitted] - Network accessline 43
def get_json(path): url = f"{API}{path}" with urllib.request.urlopen(url) as resp: - Network accessline 44
url = f"{API}{path}" with urllib.request.urlopen(url) as resp: return json.loads(resp.read()) - Network accessline 45
with urllib.request.urlopen(url) as resp: - Network accessline 50
url = f"{API}{path}" data = json.dumps(body).encode("utf-8") req = urllib.request.Request(url, data=data, headers={"Content-Type": "application/json"}, method="POST") - Network accessline 51
data = json.dumps(body).encode("utf-8") req = urllib.request.Request(url, data=data, headers={"Content-Type": "application/json"}, method="POST") with urllib.request.urlopen(req) as resp: - Network accessline 52
req = urllib.request.Request(url, data=data, headers={"Content-Type": "application/json"}, method="POST") - Network accessline 53
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_02/scripts/query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 16
[Credential-related source omitted] - Access to secretsline 16
[Credential-related source omitted] - Access to secretsline 17
[Credential-related source omitted] - Network accessline 24
if body is not None: data = json.dumps(body).encode("utf-8") req = urllib.request.Request(url, data=data, method=method) - Network accessline 25
data = json.dumps(body).encode("utf-8") req = urllib.request.Request(url, data=data, method=method) req.add_header("Content-Type", "application/json") - Network accessline 26
req = urllib.request.Request(url, data=data, method=method) - Network accessline 29
req.add_header("X-Env-Token", TOKEN) try: with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 30
try: with urllib.request.urlopen(req, timeout=30) as resp: raw = resp.read().decode("utf-8") - Network accessline 31
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 24
- `BASE_URL`: the root URL (e.g. `http://task-env:9007`) - Network accessline 28
Verify connectivity with `curl -s "$BASE_URL/health"`. Set these as shell - Network accessline 47
All GET endpoints return JSON arrays or objects. Use `curl -s` and pipe through - Network accessline 312
curl -s -X POST "$BASE_URL/api/query" \ - Network accessline 326
- [ ] `environment_access.md` was read and BASE_URL / TOKEN_HEADER extracted - [ ] All REST calls use `curl -s` with appropriate query parameters - Network accessline 327
- [ ] All REST calls use `curl -s` with appropriate query parameters
experiments/codex_skill_creator_comparison_deepseek_v4_pro_max/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_02/api_reference.md
- Network accessline 113
## Common cURL Patterns - Network accessline 116
[Credential-related source omitted] - Network accessline 117
[Credential-related source omitted] - Network accessline 119
[Credential-related source omitted] - Network accessline 120
[Credential-related source omitted] - Network accessline 122
[Credential-related source omitted] - Network accessline 123
[Credential-related source omitted] - Network accessline 125
[Credential-related source omitted] - Network accessline 126
[Credential-related source omitted] - Network accessline 131
[Credential-related source omitted] - Network accessline 132
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/cc/fewshot_attempt_01/scripts/fetch_harborcrm_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 48
url = build_url(base_url, endpoint) try: with urllib.request.urlopen(url, timeout=timeout) as response: - Network accessline 49
try: with urllib.request.urlopen(url, timeout=timeout) as response: raw = response.read().decode("utf-8") - Network accessline 50
with urllib.request.urlopen(url, timeout=timeout) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/codex/fewshot_attempt_01/scripts/harborcrm_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/deepagents/fewshot_attempt_02/scripts/fetch_harborcrm.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_01/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 22
BASE="${TASK_ENV_BASE_URL%/}" curl -sS "$BASE/api/policies" - Network accessline 23
curl -sS "$BASE/api/policies" - Network accessline 24
curl -sS "$BASE/api/crm/accounts" - Network accessline 25
curl -sS "$BASE/api/crm/contacts" - Network accessline 26
curl -sS "$BASE/api/crm/opportunities" - Network accessline 27
curl -sS "$BASE/api/crm/campaign_members"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_02/scripts/harborcrm_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 134
base = base_url.rstrip("/") target = path if path.startswith("http://") or path.startswith("https://") else f"{base}/{path.lstrip('/')}" request = urllib.request.Request(target, headers={"Accept": "application/json"}) - Network accessline 135
target = path if path.startswith("http://") or path.startswith("https://") else f"{base}/{path.lstrip('/')}" request = urllib.request.Request(target, headers={"Accept": "application/json"}) try: - Network accessline 136
request = urllib.request.Request(target, headers={"Accept": "application/json"}) - Network accessline 137
try: with urllib.request.urlopen(request, timeout=20) as response: return json.load(response) - Network accessline 138
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_001/fewshot/opencode/fewshot_attempt_03/scripts/harborcrm_utils.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 47
base = base_url.rstrip("/") + "/" url = urllib.parse.urljoin(base, path.lstrip("/")) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 48
url = urllib.parse.urljoin(base, path.lstrip("/")) with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 49
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_01/scripts/collect_medbridge.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 40
def fetch_json(base_url: str, path: str) -> Any: url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 41
url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 42
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 43
try: with urllib.request.urlopen(req, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 44
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_02/scripts/check_answer_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/cc/fewshot_attempt_03/scripts/collect_medbridge.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 42
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 43
try: with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")), None - Network accessline 44
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 16
Use [references/medbridge_rules.md](references/medbridge_rules.md) for calculation and status rules. Use [scripts/medbridge_fetch.py](scripts/medbridge_fetch.py) to gather linked records quickly without hand-writing repeated curl calls.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_01/scripts/medbridge_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_02/scripts/medbridge_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 60
description="Fetch records from the MedBridge Sales Ops task API.", ) parser.add_argument("base_url", help="Task API base URL, for example http://host:port/") - Network accessline 61
) parser.add_argument("base_url", help="Task API base URL, for example http://host:port/") parser.add_argument("--api", action="store_true", help="Fetch GET /api metadata") - Network accessline 62
parser.add_argument("base_url", help="Task API base URL, for example http://host:port/") parser.add_argument("--api", action="store_true", help="Fetch GET /api metadata") parser.add_argument(
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/codex/fewshot_attempt_03/scripts/collect_medbridge_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 39
path = "/" + path url = base + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
url = base + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 41
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 96
def main() -> int: parser = argparse.ArgumentParser(description="Collect related MedBridge Sales Ops API records.") parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/") - Network accessline 97
parser = argparse.ArgumentParser(description="Collect related MedBridge Sales Ops API records.") parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/") parser.add_argument("--qu - Network accessline 98
parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Network accessline 26
curl -sS "$BASE_URL/api/quotes/$QUOTE_ID" - Network accessline 27
curl -sS "$BASE_URL/api/products/$PRODUCT_CODE" - Network accessline 28
curl -sS "$BASE_URL/api/customers/$CUSTOMER_ID" - Network accessline 34
curl -sS "$BASE_URL/api/search?q=$QUOTE_ID" - Network accessline 35
curl -sS "$BASE_URL/api/search?q=$OPPORTUNITY_ID" - Network accessline 36
curl -sS "$BASE_URL/api/search?q=$EVENT_ID"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_01/scripts/collect_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
def get_json(base_url: str, path: str) -> object: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 42
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 43
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_02/scripts/medbridge_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/deepagents/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_02/scripts/medbridge_fetch.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/naive/fewshot_attempt_03/tools/fetch_medbridge.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
path = "/" + path url = normalize_base_url(base_url) + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
url = normalize_base_url(base_url) + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 43
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 44
try: with urllib.request.urlopen(request, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 45
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_002/fewshot/opencode/fewshot_attempt_03/scripts/collect_medbridge.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 95
def main() -> int: parser = argparse.ArgumentParser(description="Collect MedBridge Sales Ops API records.") parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/") - Network accessline 96
parser = argparse.ArgumentParser(description="Collect MedBridge Sales Ops API records.") parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/") parser.add_argument("--term", act - Network accessline 97
parser.add_argument("--base-url", required=True, help="Task API base URL, for example http://task-env:9002/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/cc/fewshot_attempt_02/scripts/collect_support_console.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 33
def get_json(base_url: str, path: str) -> Any | None: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 35
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 37
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 262
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/codex/fewshot_attempt_03/scripts/fetch_support_console.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 15
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) try: with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 16
try: with urllib.request.urlopen(url, timeout=15) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 17
with urllib.request.urlopen(url, timeout=15) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/naive/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_003/fewshot/opencode/fewshot_attempt_03/scripts/collect_support_console_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 32
url = f"{self.base_url}{path}" try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 33
try: with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 34
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 251
def main() -> int: parser = argparse.ArgumentParser(description="Collect support-console evidence for a task payload directory.") parser.add_argument("--base-url", required=True, help="Support console base URL, for example http://ta - Network accessline 252
parser = argparse.ArgumentParser(description="Collect support-console evidence for a task payload directory.") parser.add_argument("--base-url", required=True, help="Support console base URL, for example http://task-env:9003") parse - Network accessline 253
parser.add_argument("--base-url", required=True, help="Support console base URL, for example http://task-env:9003")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/cc/fewshot_attempt_03/scripts/apexcloud_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 20
BASE_URL_DEFAULT = "http://task-env:9004" - Network accessline 23
def fetch_json(url: str): with urllib.request.urlopen(url) as resp: - Network accessline 24
with urllib.request.urlopen(url) as resp: - Network accessline 28
def fetch_text(url: str) -> str: with urllib.request.urlopen(url) as resp: - Network accessline 29
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_01/scripts/apexcloud_report.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
url = base_url.rstrip("/") + path try: with urllib.request.urlopen(url) as response: - Network accessline 24
try: with urllib.request.urlopen(url) as response: return response.read().decode("utf-8") - Network accessline 25
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_01/yaml.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 9
1. Read the prompt and answer template first. Treat the template as the schema contract. 2. Identify the report family, then follow [report-playbook.md](references/report-playbook.md). 3. Use the task environment from `environment_access.md - Network accessline 10
2. Identify the report family, then follow [report-playbook.md](references/report-playbook.md). 3. Use the task environment from `environment_access.md` (`http://task-env:9004/`) for live data. 4. Return JSON only. Preserve key order, list - Network accessline 11
3. Use the task environment from `environment_access.md` (`http://task-env:9004/`) for live data.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/deepagents/fewshot_attempt_02/scripts/apex_retention_facts.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 25
def _get_json(url: str) -> Any: with urllib.request.urlopen(url) as resp: - Network accessline 26
with urllib.request.urlopen(url) as resp: - Network accessline 30
def _get_text(url: str) -> str: with urllib.request.urlopen(url) as resp: - Network accessline 31
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/naive/fewshot_attempt_03/scripts/apexcloud_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 105
def get_json(self, path, **params): with urllib.request.urlopen(self.url(path, params), timeout=30) as response: - Network accessline 106
with urllib.request.urlopen(self.url(path, params), timeout=30) as response: - Network accessline 109
def get_csv(self, path): with urllib.request.urlopen(self.url(path), timeout=30) as response: - Network accessline 110
with urllib.request.urlopen(self.url(path), timeout=30) as response: - Network accessline 663
def build_parser(): parser = argparse.ArgumentParser() parser.add_argument("--base", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9004")) - Network accessline 664
parser = argparse.ArgumentParser() parser.add_argument("--base", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9004")) sub = parser.add_subparsers(dest="command", required=True) - Network accessline 665
[Credential-related source omitted] - Access to secretsline 665
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_01/scripts/apexcloud_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 50
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: - Network accessline 51
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: return json.load(response) - Network accessline 52
with urllib.request.urlopen(url) as response: - Network accessline 56
def fetch_csv(base_url: str, path: str) -> list[dict[str, str]]: url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: - Network accessline 57
url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: text = response.read().decode("utf-8") - Network accessline 58
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 18
python scripts/apex_ops.py --prompt /path/to/input/prompt.txt --template /path/to/input/payloads/answer_template.json --base-url http://task-env:9004/ - Network accessline 20
--template /path/to/input/payloads/answer_template.json \ --base-url http://task-env:9004/ - Network accessline 21
--base-url http://task-env:9004/
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_004/fewshot/opencode/fewshot_attempt_03/scripts/apex_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 85
url = self.base + path try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 86
try: with urllib.request.urlopen(url, timeout=20) as response: data = json.loads(response.read().decode("utf-8")) - Network accessline 87
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 98
if path in self._cache: return self._cache[path] with urllib.request.urlopen(self.base + path, timeout=20) as response: - Network accessline 99
return self._cache[path] with urllib.request.urlopen(self.base + path, timeout=20) as response: text = response.read().decode("utf-8") - Network accessline 100
with urllib.request.urlopen(self.base + path, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_01/scripts/api_fetch.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/cc/fewshot_attempt_02/scripts/collect_task_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 19
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 20
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 21
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 22
Use [scripts/fetch_task_api.py](scripts/fetch_task_api.py) to fetch and paginate task API records without rewriting curl loops:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_01/scripts/fetch_task_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 38
def get_json(url): request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 39
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
try: with urllib.request.urlopen(request, timeout=20) as response: raw = response.read().decode("utf-8") - Network accessline 41
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 98
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_02/scripts/finance_api_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 67
query = urllib.parse.urlencode(params or {}) url = self.base_url + path + (("?" + query) if query else "") with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 68
url = self.base_url + path + (("?" + query) if query else "") with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 69
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/codex/fewshot_attempt_03/scripts/erp_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
if clean: url += "?" + urllib.parse.urlencode(clean) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 37
url += "?" + urllib.parse.urlencode(clean) with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 38
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_01/scripts/task_api.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/deepagents/fewshot_attempt_02/scripts/api_collect.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 17
if query: url += "?" + urllib.parse.urlencode(query) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 18
url += "?" + urllib.parse.urlencode(query) with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 19
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 25
curl -sS "$TASK_ENV_BASE_URL/api/claims?claim_id=$CLAIM_ID" - Network accessline 26
curl -sS "$TASK_ENV_BASE_URL/api/ap/bills?claim_id=$CLAIM_ID" - Network accessline 27
curl -sS "$TASK_ENV_BASE_URL/api/ap/payments?bill_id=$BILL_ID" - Network accessline 28
curl -sS "$TASK_ENV_BASE_URL/api/vendors?vendor_id=$VENDOR_ID" - Network accessline 29
curl -sS "$TASK_ENV_BASE_URL/api/compliance/objects?business_id=$BUSINESS_ID" - Network accessline 30
curl -sS "$TASK_ENV_BASE_URL/api/compliance/ownership/$BUSINESS_ID" - Network accessline 31
curl -sS "$TASK_ENV_BASE_URL/api/compliance/registry/$BUSINESS_ID" - Network accessline 32
curl -sS "$TASK_ENV_BASE_URL/api/compliance/screening/$BUSINESS_ID" - Network accessline 33
curl -sS "$TASK_ENV_BASE_URL/api/compliance/bank/$BUSINESS_ID" - Network accessline 34
curl -sS "$TASK_ENV_BASE_URL/api/prepaids/invoices?prepaid_invoice_id=$INVOICE_ID" - Network accessline 35
curl -sS "$TASK_ENV_BASE_URL/api/prepaids/gl-balances?period=$PERIOD" - Network accessline 36
curl -sS "$TASK_ENV_BASE_URL/api/close/logs"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_01/scripts/fetch_task_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
if query: url += "?" + urllib.parse.urlencode(query) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
url += "?" + urllib.parse.urlencode(query) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 33
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 35
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_005/fewshot/opencode/fewshot_attempt_03/scripts/fetch_task_group_005.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 46
def fetch_json(base_url: str, path: str, params: dict[str, Any] | None = None) -> Any: url = make_url(base_url, path, params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 47
url = make_url(base_url, path, params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(req, timeout=20) as response: - Network accessline 48
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 49
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/codex/fewshot_attempt_03/scripts/fetch_procureops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 99
def main() -> int: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Task ProcureOps base URL, for example http://host:9006/") - Network accessline 100
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Task ProcureOps base URL, for example http://host:9006/") parser.add_argument("--out", help="Write the full fetched data to this JSON path") - Network accessline 101
parser.add_argument("base_url", help="Task ProcureOps base URL, for example http://host:9006/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_02/scripts/fetch_procureops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 38
def fetch_json(base_url: str, path: str) -> object: url = base_url.rstrip("/") + "/" + path.lstrip("/") request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 39
url = base_url.rstrip("/") + "/" + path.lstrip("/") request = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 40
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 41
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_procureops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 34
def fetch_json(base_url: str, endpoint: str) -> object: url = base_url.rstrip("/") + endpoint with urllib.request.urlopen(url) as response: - Network accessline 35
url = base_url.rstrip("/") + endpoint with urllib.request.urlopen(url) as response: return json.load(response) - Network accessline 36
with urllib.request.urlopen(url) as response: - Access to secretsline 53
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 21
suppliers items programs contracts purchase_requisitions purchase_orders receipts \ ap/invoices ap/payments approvals budget_snapshots vendor_risk_events; do curl -sS "$BASE/$path" - Network accessline 22
ap/invoices ap/payments approvals budget_snapshots vendor_risk_events; do curl -sS "$BASE/$path" done - Network accessline 23
curl -sS "$BASE/$path"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/naive/fewshot_attempt_03/scripts/fetch_procureops_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
def fetch_json(base_url: str, endpoint: str, timeout: float) -> object: url = f"{base_url.rstrip('/')}/{endpoint}" request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
url = f"{base_url.rstrip('/')}/{endpoint}" request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 33
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
try: with urllib.request.urlopen(request, timeout=timeout) as response: payload = response.read().decode("utf-8") - Network accessline 35
with urllib.request.urlopen(request, timeout=timeout) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_006/fewshot/opencode/fewshot_attempt_02/scripts/fetch_procureops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 35
def fetch_json(url: str, timeout: float) -> object: request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 37
with urllib.request.urlopen(request, timeout=timeout) as response: - Network accessline 42
def main() -> int: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="ProcureOps API base URL, for example http://task-env:9006/") - Network accessline 43
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="ProcureOps API base URL, for example http://task-env:9006/") parser.add_argument( - Network accessline 44
parser.add_argument("base_url", help="ProcureOps API base URL, for example http://task-env:9006/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_02/scripts/fetch_erp.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 67
base = base_url.rstrip("/") url = base + validate_path(path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 68
url = base + validate_path(path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 69
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 70
try: with urllib.request.urlopen(request, timeout=timeout) as response: body = response.read().decode("utf-8") - Network accessline 71
with urllib.request.urlopen(request, timeout=timeout) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/cc/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_02/scripts/northwind_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
query = urllib.parse.urlencode(params) url = f"{url}?{query}" with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 32
url = f"{url}?{query}" with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 33
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/deepagents/fewshot_attempt_03/scripts/northwind_erp.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/naive/fewshot_attempt_03/scripts/northwind_erp.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 22
if params: query = "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(base_url + path + query) as response: - Network accessline 23
query = "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(base_url + path + query) as response: return json.load(response) - Network accessline 24
with urllib.request.urlopen(base_url + path + query) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_02/references/northwind-erp.md
- Network accessline 5
Base URL: `http://task-env:9007/`
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_007/fewshot/opencode/fewshot_attempt_03/scripts/northwind_erp.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 13
[Credential-related source omitted] - Access to secretsline 13
[Credential-related source omitted] - Network accessline 21
if params: url = f"{url}?{urllib.parse.urlencode(params)}" with urllib.request.urlopen(url) as resp: - Network accessline 22
url = f"{url}?{urllib.parse.urlencode(params)}" with urllib.request.urlopen(url) as resp: return json.load(resp) - Network accessline 23
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_02/scripts/advisory_planner.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 634
def main() -> int: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/")) - Network accessline 635
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/")) parser.add_argument("--client-id") - Network accessline 636
[Credential-related source omitted] - Access to secretsline 636
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/cc/fewshot_attempt_03/scripts/solve_advisory_task.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 37
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=30) as response: return json.load(response) - Network accessline 38
with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 597
parser.add_argument( "--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/"), - Network accessline 598
"--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/"), help="Advisory API base URL.", - Network accessline 599
[Credential-related source omitted] - Access to secretsline 599
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_01/references/environment.md
- Network accessline 3
Use the harness-provided `API_BASE` when present. The staged fallback base URL is `http://task-env:9008/`.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 21
API_BASE="${API_BASE:-http://task-env:9008/}" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_02/scripts/advisory_planning_solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 80
if params: url += "?" + urllib.parse.urlencode(params) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 81
url += "?" + urllib.parse.urlencode(params) request = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(request, timeout=15) as response: - Network accessline 82
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 83
with urllib.request.urlopen(request, timeout=15) as response: - Network accessline 571
task_id = infer_task_id(input_dir, args.task_id) client_id = infer_client_id(prompt, memo) api_base = args.api_base or os.environ.get("API_BASE") or "http://task-env:9008/" - Network accessline 572
client_id = infer_client_id(prompt, memo) api_base = args.api_base or os.environ.get("API_BASE") or "http://task-env:9008/" api = AdvisoryApi(api_base) - Network accessline 573
[Credential-related source omitted] - Access to secretsline 573
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/codex/fewshot_attempt_03/scripts/advisory_planning.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 132
if not task_id: raise SystemExit("Could not infer task_id; pass --task-id") api_base = args.api_base or os.environ.get("API_BASE") or "http://task-env:9008/" - Network accessline 134
[Credential-related source omitted] - Access to secretsline 134
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_02/scripts/advisory_planner.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 85
[Credential-related source omitted] - Network accessline 595
def main() -> int: parser = argparse.ArgumentParser(description="Build advisory planning JSON from task API records.") parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008")) - Network accessline 596
parser = argparse.ArgumentParser(description="Build advisory planning JSON from task API records.") parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008")) parser.add_argument("--client-id") - Network accessline 597
[Credential-related source omitted] - Access to secretsline 597
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/deepagents/fewshot_attempt_03/scripts/solve_advisory_task.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 27
def fetch_json(api_base: str, path: str) -> Any: url = api_base.rstrip("/") + path with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 28
url = api_base.rstrip("/") + path with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 29
with urllib.request.urlopen(url, timeout=20) as response: - Access to secretsline 461
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/naive/fewshot_attempt_01/solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
def fetch_json(api_base: str, path: str) -> Any: url = api_base.rstrip("/") + path with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 38
url = api_base.rstrip("/") + path with urllib.request.urlopen(url, timeout=15) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 575
def main() -> int: parser = argparse.ArgumentParser() parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/")) - Network accessline 576
parser = argparse.ArgumentParser() parser.add_argument("--api-base", default=os.environ.get("API_BASE", "http://task-env:9008/")) parser.add_argument("--prompt") - Network accessline 577
[Credential-related source omitted] - Access to secretsline 577
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_008/fewshot/opencode/fewshot_attempt_03/scripts/fetch_case_data.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_02/scripts/solve_finance_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 84
[Credential-related source omitted] - Network accessline 99
def fetch_json(base_url: str, endpoint: str): url = base_url.rstrip("/") + endpoint with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 100
url = base_url.rstrip("/") + endpoint with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 101
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/cc/fewshot_attempt_03/scripts/solve_crescent_ops.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_02/SKILL.md
- Network accessline 22
python /path/to/skill/scripts/finance_ops_report.py /path/to/task/input --base-url http://task-env:9009/
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/codex/fewshot_attempt_02/scripts/finance_ops_report.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 78
[Credential-related source omitted] - Network accessline 86
return "http://task-env:9009" - Network accessline 96
url = f"{self.base_url}{endpoint}" try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 97
try: with urllib.request.urlopen(url, timeout=20) as response: self.cache[endpoint] = json.load(response) - Network accessline 98
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_01/scripts/crescent_finance_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 61
[Credential-related source omitted] - Access to secretsline 62
[Credential-related source omitted] - Access to secretsline 63
[Credential-related source omitted] - Network accessline 80
def fetch(base_url: str, path: str): - Network accessline 81
with urllib.request.urlopen(base_url + path) as response: - Network accessline 105
class Finance: def __init__(self, base_url: str): self.branches = fetch(base_url, "/api/finance/branches") - Network accessline 106
def __init__(self, base_url: str): self.branches = fetch(base_url, "/api/finance/branches") self.period_map = fetch(base_url, "/api/finance/period-map") - Network accessline 107
self.branches = fetch(base_url, "/api/finance/branches") - Network accessline 108
self.period_map = fetch(base_url, "/api/finance/period-map") - Network accessline 109
self.accounts = fetch(base_url, "/api/finance/accounts") - Network accessline 110
self.records = fetch(base_url, "/api/finance/records") - Network accessline 365
class Compensation: def __init__(self, base_url: str): self.rate_book = fetch(base_url, "/api/compensation/rate-book") - Network accessline 366
def __init__(self, base_url: str): self.rate_book = fetch(base_url, "/api/compensation/rate-book") self.rosters = fetch(base_url, "/api/compensation/rosters") - Network accessline 367
self.rate_book = fetch(base_url, "/api/compensation/rate-book") - Network accessline 368
self.rosters = fetch(base_url, "/api/compensation/rosters") - Network accessline 369
self.scenarios = fetch(base_url, "/api/compensation/scenarios") - Network accessline 539
class Payroll: def __init__(self, base_url: str): self.rate_book = fetch(base_url, "/api/payroll/rate-book") - Network accessline 540
def __init__(self, base_url: str): self.rate_book = fetch(base_url, "/api/payroll/rate-book") self.productions = fetch(base_url, "/api/payroll/productions") - Network accessline 541
self.rate_book = fetch(base_url, "/api/payroll/rate-book") - Network accessline 542
self.productions = fetch(base_url, "/api/payroll/productions")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/deepagents/fewshot_attempt_03/scripts/finance_ops_solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 54
[Credential-related source omitted] - Network accessline 56
base = payload_base if not base: base = "http://task-env:9009/" - Network accessline 57
if not base: base = "http://task-env:9009/" return base.rstrip("/") - Network accessline 58
base = "http://task-env:9009/" - Network accessline 62
def fetch_json(base_url: str, endpoint: str) -> Any: url = f"{base_url}{endpoint}" with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 63
url = f"{base_url}{endpoint}" with urllib.request.urlopen(url, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 64
with urllib.request.urlopen(url, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 22
It prints one JSON object. Before returning it, check that the top-level keys and list ordering match `answer_template.json`. If the payload base URL is a placeholder, the script falls back to `http://task-env:9009/`.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_01/scripts/solve_finance_ops.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
def fetch_json(base_url: str, path: str): url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: - Network accessline 31
url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: return json.loads(response.read().decode("utf-8"), parse_float=Decimal) - Network accessline 32
with urllib.request.urlopen(url) as response: - Network accessline 639
base_url = environment_payload.get("base_url") or "" if base_url == "<TASK_ENV_BASE_URL>" or not base_url: base_url = os.environ.get("TASK_ENV_BASE_URL") or os.environ.get("FINANCE_OPS_BASE_URL") or "http://task-env:9009/" - Network accessline 640
if base_url == "<TASK_ENV_BASE_URL>" or not base_url: base_url = os.environ.get("TASK_ENV_BASE_URL") or os.environ.get("FINANCE_OPS_BASE_URL") or "http://task-env:9009/" return base_url.rstrip("/") - Network accessline 641
[Credential-related source omitted] - Access to secretsline 641
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_02/crescent_ops_solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
url = base_url.rstrip("/") + endpoint try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 38
try: with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 39
with urllib.request.urlopen(url, timeout=20) as response: - Access to secretsline 48
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_009/fewshot/naive/fewshot_attempt_03/finance_ops_solver.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 71
def get(self, endpoint: str) -> Any: url = f"{self.base_url}{endpoint}" with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 72
url = f"{self.base_url}{endpoint}" with urllib.request.urlopen(url, timeout=20) as response: return json.load(response) - Network accessline 73
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 96
[Credential-related source omitted] - Access to secretsline 96
[Credential-related source omitted] - Network accessline 97
[Credential-related source omitted] - Access to secretsline 97
[Credential-related source omitted] - Network accessline 98
or "http://task-env:9009/"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/cc/fewshot_attempt_01/references/environment.md
- Network accessline 3
Base URL: `http://task-env:9010`
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 12
1. Read the prompt, `input/payloads/answer_template.json`, and every local payload. 2. Fetch current records from the environment. Default base URL: `http://task-env:9010/`. - Network accessline 13
2. Fetch current records from the environment. Default base URL: `http://task-env:9010/`.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_02/scripts/asteria_tools.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 16
DEFAULT_BASE_URL = "http://task-env:9010" - Network accessline 21
url = base_url.rstrip("/") + "/" + path.lstrip("/") try: with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 22
try: with urllib.request.urlopen(url, timeout=15) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 23
with urllib.request.urlopen(url, timeout=15) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/deepagents/fewshot_attempt_03/scripts/asteria_env_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 68
def parse_access(path: Path | None) -> tuple[str, set[str]]: base_url = "http://task-env:9010/" - Network accessline 69
base_url = "http://task-env:9010/" - Network accessline 99
def fetch(base_url: str, path: str, allowed: set[str]) -> Any | None: - Network accessline 407
"policies": "/api/policies", } env = {name: fetch(base_url, path, allowed) for name, path in endpoints.items()} - Network accessline 409
env = {name: fetch(base_url, path, allowed) for name, path in endpoints.items()} - Network accessline 419
for portfolio_id in request["portfolio_ids"]: path = f"/api/portfolios/{portfolio_id}/holdings" holdings_response = fetch(base_url, path, allowed) - Network accessline 420
path = f"/api/portfolios/{portfolio_id}/holdings" holdings_response = fetch(base_url, path, allowed) if not isinstance(holdings_response, dict): - Network accessline 421
holdings_response = fetch(base_url, path, allowed)
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_01/scripts/asteria_calcs.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 15
def fetch_json(base_url: str, path: str) -> Any: base = base_url.rstrip("/") with urllib.request.urlopen(f"{base}{path}", timeout=20) as response: - Network accessline 16
base = base_url.rstrip("/") with urllib.request.urlopen(f"{base}{path}", timeout=20) as response: return json.load(response) - Network accessline 17
with urllib.request.urlopen(f"{base}{path}", timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 20
Default base URL is `http://task-env:9010/` when no local access note says otherwise.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_02/scripts/asteria_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 17
DEFAULT_BASE_URL = "http://task-env:9010/" - Network accessline 29
url = normalize_base_url(base_url) + path.lstrip("/") try: with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 30
try: with urllib.request.urlopen(url, timeout=15) as response: return json.load(response) - Network accessline 31
with urllib.request.urlopen(url, timeout=15) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_03/SKILL.md
- Network accessline 12
1. Read `prompt.txt`, every JSON file in `input/payloads/`, and especially `answer_template.json`. 2. Fetch current data from `http://task-env:9010/` unless the task gives another base URL. Use only read-only endpoints. Never call a judge o - Network accessline 13
2. Fetch current data from `http://task-env:9010/` unless the task gives another base URL. Use only read-only endpoints. Never call a judge or write endpoint.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/naive/fewshot_attempt_03/asteria_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 24
[Credential-related source omitted] - Access to secretsline 24
[Credential-related source omitted] - Network accessline 28
url = f"{BASE_URL}{path if path.startswith('/') else '/' + path}" try: with urllib.request.urlopen(url, timeout=10) as response: - Network accessline 29
try: with urllib.request.urlopen(url, timeout=10) as response: return json.load(response) - Network accessline 30
with urllib.request.urlopen(url, timeout=10) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_01/scripts/asteria_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 72
url = base_url.rstrip("/") + path try: with urllib.request.urlopen(url, timeout=10) as resp: - Network accessline 73
try: with urllib.request.urlopen(url, timeout=10) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 74
with urllib.request.urlopen(url, timeout=10) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_02/scripts/asteria_toolkit.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_010/fewshot/opencode/fewshot_attempt_03/scripts/asteria_tools.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 60
def fetch_json(url: str) -> Any: req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 61
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 62
try: with urllib.request.urlopen(req, timeout=30) as resp: raw = resp.read().decode("utf-8") - Network accessline 63
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_02/scripts/credit_office_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
url = f"{base}{path}" try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 24
try: with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 25
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 1001
def build_parser() -> argparse.ArgumentParser: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task API base URL, e.g. http://task-env:9011") - Network accessline 1002
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task API base URL, e.g. http://task-env:9011") subparsers = parser.add_subparsers(dest="command", required=True) - Network accessline 1003
parser.add_argument("--base-url", required=True, help="Task API base URL, e.g. http://task-env:9011")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/cc/fewshot_attempt_03/scripts/credit_office_tools.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 53
url = urllib.parse.urljoin(base, path.lstrip("/")) try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 54
try: with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 55
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 60
- Fetch the segment endpoint, the NC row from the NCUA benchmark table, and the policy.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/codex/fewshot_attempt_03/scripts/case_pack.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 13
def fetch_json(base_url: str, path: str): url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: - Network accessline 14
url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: return json.load(response) - Network accessline 15
with urllib.request.urlopen(url) as response: - Network accessline 28
def main() -> int: parser = argparse.ArgumentParser(description="Fetch a compact credit-office case bundle.") parser.add_argument("--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9011")) - Network accessline 29
parser = argparse.ArgumentParser(description="Fetch a compact credit-office case bundle.") parser.add_argument("--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9011")) parser.add_argument("--branch", help=" - Network accessline 30
[Credential-related source omitted] - Access to secretsline 30
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_02/scripts/credit_office_toolkit.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/deepagents/fewshot_attempt_03/scripts/credit_committee.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 42
def get(self, path: str) -> Any: url = urllib.parse.urljoin(self.base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 43
url = urllib.parse.urljoin(self.base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 44
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 45
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 46
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 908
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_011/fewshot/opencode/fewshot_attempt_03/scripts/credit_review_toolkit.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 29
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(normalize_base_url(base_url), path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 30
url = urllib.parse.urljoin(normalize_base_url(base_url), path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 31
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
try: with urllib.request.urlopen(request, timeout=15) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 33
with urllib.request.urlopen(request, timeout=15) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/cc/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/codex/fewshot_attempt_03/scripts/collect_peopleops_evidence.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_01/scripts/collect_peopleops_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 14
[Credential-related source omitted] - Access to secretsline 14
[Credential-related source omitted] - Network accessline 17
def fetch_json(base_url: str, path: str) -> Any: url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: - Network accessline 18
url = base_url.rstrip("/") + path with urllib.request.urlopen(url) as response: return json.load(response) - Network accessline 19
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_02/scripts/peopleops_collect.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/deepagents/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_02/scripts/collect_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
def fetch_text(base_url: str, path: str) -> str: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 31
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) with urllib.request.urlopen(url, timeout=15) as response: return response.read().decode("utf-8") - Network accessline 32
with urllib.request.urlopen(url, timeout=15) as response: - Network accessline 159
def parse_args(argv: list[str]) -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Task environment base URL, for example http://task-env:9012/") - Network accessline 160
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Task environment base URL, for example http://task-env:9012/") parser.add_argument("--employee-id", action="append", default=[], help="Emplo - Network accessline 161
parser.add_argument("base_url", help="Task environment base URL, for example http://task-env:9012/") parser.add_argument("--employee-id", action="append", default=[], help="Employee id to filter for") parser.add_argument("--case-id"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/naive/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_02/scripts/fetch_peopleops_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
def get_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 33
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 74
def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task environment base URL, e.g. http://task-env:9012/") - Network accessline 75
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task environment base URL, e.g. http://task-env:9012/") parser.add_argument("--query", action="append", default=[], help="C - Network accessline 76
parser.add_argument("--base-url", required=True, help="Task environment base URL, e.g. http://task-env:9012/") parser.add_argument("--query", action="append", default=[], help="Case, employee, candidate, opening, audit, or record id/nam
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03/scripts/fetch_peopleops_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 35
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 37
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 38
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 140
def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="PeopleOps task base URL, for example http://task-env:9012/") - Network accessline 141
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="PeopleOps task base URL, for example http://task-env:9012/") parser.add_argument("--out", default="-", help="Output JSON path, or '-' for st - Network accessline 142
parser.add_argument("base_url", help="PeopleOps task base URL, for example http://task-env:9012/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_012/fewshot/opencode/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/cc/fewshot_attempt_02/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method="POST" if payload is not None else "GET") - Network accessline 24
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method="POST" if payload is not None else "GET") try: - Network accessline 25
req = urllib.request.Request(url, data=data, headers=headers, method="POST" if payload is not None else "GET") - Network accessline 26
try: with urllib.request.urlopen(req, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 27
with urllib.request.urlopen(req, timeout=30) as response: - Network accessline 228
def build_parser() -> argparse.ArgumentParser: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task environment base URL, for example http://task-env:9013") - Network accessline 229
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Task environment base URL, for example http://task-env:9013") sub = parser.add_subparsers(dest="command", required=True) - Network accessline 230
parser.add_argument("--base-url", required=True, help="Task environment base URL, for example http://task-env:9013")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_01/scripts/fetch_portal_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers) - Network accessline 31
headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers) with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 32
request = urllib.request.Request(url, data=data, headers=headers) - Network accessline 33
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_02/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 32
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 33
headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 34
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 35
try: with urllib.request.urlopen(request, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 36
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_03/scripts/check_json_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/codex/fewshot_attempt_03/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 25
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 26
headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 27
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 28
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 29
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_01/scripts/portal_collect.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
def get_json(base_url: str, path: str) -> object: url = normalize_base(base_url) + path with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 38
url = normalize_base(base_url) + path with urllib.request.urlopen(url, timeout=30) as response: return json.load(response) - Network accessline 39
with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 44
url = normalize_base(base_url) + "/query" body = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( - Network accessline 45
body = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( url, - Network accessline 46
request = urllib.request.Request( - Network accessline 50
method="POST", ) with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 51
) with urllib.request.urlopen(request, timeout=30) as response: return json.load(response) - Network accessline 52
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_02/scripts/collect_portal_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
def _get(base_url: str, path: str) -> Any: with urllib.request.urlopen(_base(base_url) + path, timeout=20) as response: - Network accessline 24
with urllib.request.urlopen(_base(base_url) + path, timeout=20) as response: - Network accessline 28
def _post_sql(base_url: str, sql: str) -> Any: data = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( - Network accessline 29
data = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( _base(base_url) + "/query", - Network accessline 30
request = urllib.request.Request( - Network accessline 34
method="POST", ) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 35
) with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 36
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/deepagents/fewshot_attempt_03/scripts/portal_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 19
[Credential-related source omitted] - Network accessline 25
def _get(base: str, path: str) -> object: url = urllib.parse.urljoin(base, path.lstrip("/")) with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 26
url = urllib.parse.urljoin(base, path.lstrip("/")) with urllib.request.urlopen(url, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 27
with urllib.request.urlopen(url, timeout=30) as response: - Network accessline 32
url = urllib.parse.urljoin(base, path.lstrip("/")) data = json.dumps(payload).encode("utf-8") request = urllib.request.Request( - Network accessline 33
data = json.dumps(payload).encode("utf-8") request = urllib.request.Request( url, - Network accessline 34
request = urllib.request.Request( - Network accessline 38
method="POST", ) with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 39
) with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 40
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_01/SKILL.md
- Network accessline 17
curl -sS -X POST "$BASE_URL/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/naive/fewshot_attempt_03/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 15
url = base_url.rstrip("/") + "/query" data = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( - Network accessline 16
data = json.dumps({"sql": sql}).encode("utf-8") request = urllib.request.Request( url, - Network accessline 17
request = urllib.request.Request( - Network accessline 21
method="POST", ) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 22
) with urllib.request.urlopen(request, timeout=20) as response: payload = json.load(response) - Network accessline 23
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 171
def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Portal base URL, for example http://task-env:9013") - Network accessline 172
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Portal base URL, for example http://task-env:9013") parser.add_argument("--kind", required=True, choices=sorted(SNAPSHOTTER - Network accessline 173
parser.add_argument("--base-url", required=True, help="Portal base URL, for example http://task-env:9013")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 29
curl -sS "$BASE/patients/$patient_id" - Network accessline 30
curl -sS "$BASE/referrals/$referral_id" - Network accessline 31
curl -sS "$BASE/transfers/$transfer_id" - Network accessline 32
curl -sS "$BASE/chart/$patient_id" - Network accessline 33
curl -sS "$BASE/programs/$program_code/candidates" - Network accessline 34
curl -sS "$BASE/icd/$icd10_code" - Network accessline 40
curl -sS "$BASE/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_013/fewshot/opencode/fewshot_attempt_02/scripts/cedar_intake_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 105
def query(self, sql: str) -> list[dict[str, Any]]: body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( - Network accessline 106
body = json.dumps({"sql": sql}).encode("utf-8") req = urllib.request.Request( self.base_url + "query", - Network accessline 107
req = urllib.request.Request( - Network accessline 112
) try: with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 113
try: with urllib.request.urlopen(req, timeout=20) as resp: payload = json.loads(resp.read().decode("utf-8")) - Network accessline 114
with urllib.request.urlopen(req, timeout=20) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_01/scripts/northstar_sql.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02/scripts/northstar_sql.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 32
try: with urllib.request.urlopen(req, timeout=30) as response: raw = response.read().decode("utf-8") - Network accessline 33
with urllib.request.urlopen(req, timeout=30) as response: - Access to secretsline 43
[Credential-related source omitted] - Access to secretsline 44
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_02/scripts/template_check.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/cc/fewshot_attempt_03/references/northstar-payer-ops.md
- Network accessline 8
curl -sS -X POST "$BASE_URL/sql/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_01/scripts/northstar_sql.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 24
base_url = args.base_url.rstrip("/") request = urllib.request.Request( - Network accessline 25
request = urllib.request.Request( - Network accessline 35
try: with urllib.request.urlopen(request, timeout=args.timeout) as response: - Network accessline 36
with urllib.request.urlopen(request, timeout=args.timeout) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_03/references/northstar_environment.md
- Network accessline 8
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/codex/fewshot_attempt_03/scripts/northstar_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 34
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 35
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 36
with urllib.request.urlopen(request, timeout=30) as response: - Access to secretsline 59
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_01/scripts/northstar_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 28
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 29
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 30
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 31
try: with urllib.request.urlopen(req, timeout=30) as resp: raw = resp.read().decode("utf-8") - Network accessline 32
with urllib.request.urlopen(req, timeout=30) as resp: - Access to secretsline 48
[Credential-related source omitted] - Access to secretsline 49
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_02/SKILL.md
- Network accessline 21
curl -sS "$BASE_URL/sql/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/deepagents/fewshot_attempt_03/scripts/query_northstar.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL"), help="Environment base URL, for example http://task-env:9014", - Network accessline 24
[Credential-related source omitted] - Access to secretsline 24
[Credential-related source omitted] - Network accessline 25
help="Environment base URL, for example http://task-env:9014", - Access to secretsline 29
[Credential-related source omitted] - Network accessline 60
request = urllib.request.Request( - Network accessline 67
try: with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 68
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_01/references/northstar-data-map.md
- Network accessline 8
curl -sS -X POST "$BASE_URL/sql/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/naive/fewshot_attempt_02/scripts/northstar_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 28
[Credential-related source omitted] - Network accessline 29
headers["Authorization"] = f"Bearer {token}" request = urllib.request.Request(url, data=data, headers=headers, method="POST" if payload else "GET") try: - Network accessline 30
request = urllib.request.Request(url, data=data, headers=headers, method="POST" if payload else "GET") - Network accessline 31
try: with urllib.request.urlopen(request, timeout=30) as response: body = response.read() - Network accessline 32
with urllib.request.urlopen(request, timeout=30) as response: - Access to secretsline 48
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Access to secretsline 27
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_01/scripts/northstar_env.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_014/fewshot/opencode/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_01/scripts/ehr_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 27
def get_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 28
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 29
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 30
try: with urllib.request.urlopen(request, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 31
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_02/scripts/check_json_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_02/scripts/collect_ehr_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 29
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 30
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 31
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
try: with urllib.request.urlopen(req, timeout=20) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 33
with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 56
parser.add_argument( "--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9015/"), - Network accessline 57
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9015/"), help="Task environment base URL.", - Network accessline 58
[Credential-related source omitted] - Access to secretsline 58
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_03/SKILL.md
- Network accessline 2
name: ehr-quality-json description: Use this skill for EHR quality-governance tasks that require reading a read-only task environment API and returning normalized JSON packets, especially duplicate-chart merge readiness, referral coordinati - Network accessline 3
description: Use this skill for EHR quality-governance tasks that require reading a read-only task environment API and returning normalized JSON packets, especially duplicate-chart merge readiness, referral coordination, care-transition han - Network accessline 4
compatibility: Requires shell access with curl or Python 3 for HTTP GET requests to the task environment described by the prompt or environment_access.md. - Network accessline 31
The script only performs GET requests and prints a JSON evidence bundle. Manual `curl -s` calls are also fine.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/cc/fewshot_attempt_03/scripts/ehr_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 27
def get_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 28
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 29
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 30
try: with urllib.request.urlopen(req, timeout=20) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 31
with urllib.request.urlopen(req, timeout=20) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_01/scripts/fetch_ehr_context.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_02/scripts/ehr_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 39
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 41
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
try: with urllib.request.urlopen(request, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 43
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/codex/fewshot_attempt_03/scripts/fetch_ehr_evidence.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_01/scripts/ehr_quality_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 33
if params: url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 35
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
try: with urllib.request.urlopen(req, timeout=20) as resp: text = resp.read().decode("utf-8") - Network accessline 37
with urllib.request.urlopen(req, timeout=20) as resp: - Access to secretsline 140
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_ehr.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 28
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 29
try: with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 30
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 49
parser.add_argument( "--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9015/"), - Network accessline 50
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9015/"), ) - Network accessline 51
[Credential-related source omitted] - Access to secretsline 51
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/naive/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_02/scripts/ehr_collect.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 31
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 32
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 33
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 34
try: with urllib.request.urlopen(request, timeout=20) as response: return json.load(response) - Network accessline 35
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 89
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_015/fewshot/opencode/fewshot_attempt_03/scripts/ehr_case_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 39
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) try: with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 40
try: with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 41
with urllib.request.urlopen(url, timeout=20) as response: - Access to secretsline 67
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_01/scripts/clinic_protocol_helper.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_02/scripts/collect_clinic_context.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/cc/fewshot_attempt_03/scripts/fetch_runtime.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 39
def get_json(base_url: str, path: str, timeout: float) -> tuple[Any | None, dict[str, Any]]: url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) started = time.time() - Network accessline 41
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 42
started = time.time() try: with urllib.request.urlopen(request, timeout=timeout) as response: - Network accessline 43
try: with urllib.request.urlopen(request, timeout=timeout) as response: body = response.read() - Network accessline 44
with urllib.request.urlopen(request, timeout=timeout) as response: - Access to secretsline 154
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_01/scripts/check_answer_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_01/scripts/collect_clinic_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 43
def fetch_json(base_url: str, path: str) -> Any: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, method="GET") - Network accessline 44
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, method="GET") with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 45
request = urllib.request.Request(url, method="GET") - Network accessline 46
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/codex/fewshot_attempt_03/scripts/fetch_clinic_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 23
def get_json(base_url: str, path: str) -> object: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 24
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 25
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 26
try: with urllib.request.urlopen(request, timeout=20) as response: charset = response.headers.get_content_charset() or "utf-8" - Network accessline 27
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_01/scripts/fetch_clinic_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 32
def fetch_json(base_url: str, path: str, timeout: float) -> tuple[Any | None, str | None]: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "applicatio - Network accessline 33
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 34
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 35
try: with urllib.request.urlopen(request, timeout=timeout) as response: body = response.read().decode("utf-8") - Network accessline 36
with urllib.request.urlopen(request, timeout=timeout) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_02/scripts/collect_context.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/deepagents/fewshot_attempt_03/scripts/collect_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 26
def get_json(base_url: str, path: str) -> object: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, method="GET") - Network accessline 27
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, method="GET") try: - Network accessline 28
request = urllib.request.Request(url, method="GET") - Network accessline 29
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 30
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/naive/fewshot_attempt_02/scripts/collect_clinic_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 45
[Credential-related source omitted] - Network accessline 46
headers[header_name] = token request = urllib.request.Request(url, headers=headers, method="GET") try: - Network accessline 47
request = urllib.request.Request(url, headers=headers, method="GET") - Network accessline 48
try: with urllib.request.urlopen(request, timeout=15) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 49
with urllib.request.urlopen(request, timeout=15) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_01/scripts/check_answer_against_template.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_01/scripts/collect_clinic_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 32
def get_json(base_url: str, endpoint: str) -> object: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", endpoint.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 33
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", endpoint.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 34
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 35
try: with urllib.request.urlopen(request, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 36
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_02/scripts/fetch_case_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 46
def get_json(base_url: str, path: str) -> Any: url = urljoin(base_url, path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 47
url = urljoin(base_url, path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 48
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 49
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 50
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_02/scripts/validate_answer_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_03/scripts/check_json_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_016/fewshot/opencode/fewshot_attempt_03/scripts/fetch_clinic_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 4
The script intentionally uses only Python's standard library. It writes raw JSON responses so the solver can inspect source data without repeatedly retyping curl commands. It does not mutate the runtime and does not call POST endpoints. - Network accessline 5
responses so the solver can inspect source data without repeatedly retyping curl commands. It does not mutate the runtime and does not call POST endpoints. """ - Network accessline 95
def main() -> int: parser = argparse.ArgumentParser(description="Fetch standard synthetic clinic GET endpoints.") parser.add_argument("--base-url", required=True, help="Runtime base URL, for example http://task-env:9016/") - Network accessline 96
parser = argparse.ArgumentParser(description="Fetch standard synthetic clinic GET endpoints.") parser.add_argument("--base-url", required=True, help="Runtime base URL, for example http://task-env:9016/") parser.add_argument("--case- - Network accessline 97
parser.add_argument("--base-url", required=True, help="Runtime base URL, for example http://task-env:9016/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_01/scripts/fetch_review_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 58
data = json.dumps(body).encode("utf-8") req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers) - Network accessline 59
req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers) with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 60
req = urllib.request.Request(url, data=data, headers=req_headers) - Network accessline 61
with urllib.request.urlopen(req, timeout=20) as resp: - Access to secretsline 92
[Credential-related source omitted] - Access to secretsline 95
[Credential-related source omitted] - Access to secretsline 96
[Credential-related source omitted] - Access to secretsline 131
[Credential-related source omitted] - Access to secretsline 149
[Credential-related source omitted] - Network accessline 166
parser.add_argument( "--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9017/"), - Network accessline 167
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9017/"), help="Investigation Review Hub base URL", - Network accessline 168
[Credential-related source omitted] - Access to secretsline 168
[Credential-related source omitted] - Access to secretsline 173
[Credential-related source omitted] - Access to secretsline 181
[Credential-related source omitted] - Access to secretsline 182
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_02/scripts/hub_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
data = json.dumps(body).encode("utf-8") req_headers.setdefault("Content-Type", "application/json") req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 19
req_headers.setdefault("Content-Type", "application/json") req = urllib.request.Request(url, data=data, headers=req_headers, method=method) try: - Network accessline 20
req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 21
try: with urllib.request.urlopen(req, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 22
with urllib.request.urlopen(req, timeout=20) as response: - Network accessline 33
def main() -> int: parser = argparse.ArgumentParser(description="Fetch matter-scoped Hub rows for answer drafting.") parser.add_argument("--base-url", required=True, help="Investigation Review Hub base URL, for example http://task-e - Network accessline 34
parser = argparse.ArgumentParser(description="Fetch matter-scoped Hub rows for answer drafting.") parser.add_argument("--base-url", required=True, help="Investigation Review Hub base URL, for example http://task-env:9017") parser.ad - Network accessline 35
parser.add_argument("--base-url", required=True, help="Investigation Review Hub base URL, for example http://task-env:9017") - Access to secretsline 43
[Credential-related source omitted] - Access to secretsline 44
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_03/scripts/hub_dump.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 54
data = json.dumps(payload).encode("utf-8") req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 55
req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers, method=method) with urllib.request.urlopen(req, timeout=30) as response: - Network accessline 56
req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 57
with urllib.request.urlopen(req, timeout=30) as response: - Access to secretsline 170
[Credential-related source omitted] - Access to secretsline 171
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/cc/fewshot_attempt_03/scripts/validate_template_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_01/scripts/hub_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 57
[Credential-related source omitted] - Access to secretsline 63
[Credential-related source omitted] - Access to secretsline 64
[Credential-related source omitted] - Network accessline 66
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 67
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 68
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 83
[Credential-related source omitted] - Access to secretsline 94
[Credential-related source omitted] - Access to secretsline 122
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_02/scripts/collect_hub_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 163
[Credential-related source omitted] - Access to secretsline 164
[Credential-related source omitted] - Access to secretsline 180
[Credential-related source omitted] - Access to secretsline 183
[Credential-related source omitted] - Access to secretsline 192
[Credential-related source omitted] - Access to secretsline 198
[Credential-related source omitted] - Access to secretsline 199
[Credential-related source omitted] - Access to secretsline 212
[Credential-related source omitted] - Access to secretsline 216
[Credential-related source omitted] - Access to secretsline 367
[Credential-related source omitted] - Access to secretsline 380
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/codex/fewshot_attempt_03/scripts/collect_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 49
data = json.dumps(body).encode("utf-8") req_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 50
req_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=req_headers, method=method) try: - Network accessline 51
request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 52
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 53
with urllib.request.urlopen(request, timeout=30) as response: - Access to secretsline 112
[Credential-related source omitted] - Access to secretsline 113
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Access to secretsline 19
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_01/scripts/fetch_review_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 53
def request_json(url: str, headers: dict[str, str], timeout: float) -> Any: request = urllib.request.Request(url, headers=headers) - Network accessline 54
request = urllib.request.Request(url, headers=headers) - Network accessline 55
with urllib.request.urlopen(request, timeout=timeout) as response: - Network accessline 64
request_headers = dict(headers) request_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=body, headers=request_headers) - Network accessline 65
request_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=body, headers=request_headers) with urllib.request.urlopen(request, timeout=timeout) as response: - Network accessline 66
request = urllib.request.Request(url, data=body, headers=request_headers) - Network accessline 67
with urllib.request.urlopen(request, timeout=timeout) as response: - Access to secretsline 152
[Credential-related source omitted] - Access to secretsline 153
[Credential-related source omitted] - Access to secretsline 164
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_02/scripts/collect_review_hub.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Access to secretsline 17
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_hub_matter.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 31
[Credential-related source omitted] - Network accessline 36
[Credential-related source omitted] - Access to secretsline 36
[Credential-related source omitted] - Network accessline 37
[Credential-related source omitted] - Access to secretsline 37
[Credential-related source omitted] - Network accessline 38
request = urllib.request.Request(url, data=payload, headers=headers, method="POST") - Network accessline 39
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 40
with urllib.request.urlopen(request, timeout=20) as response: - Access to secretsline 59
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_01/SKILL.md
- Access to secretsline 21
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_01/scripts/hub_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 74
data = json.dumps(payload).encode("utf-8") req_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 75
req_headers["Content-Type"] = "application/json" request = urllib.request.Request(url, data=data, headers=req_headers, method=method) with urllib.request.urlopen(request, timeout=30) as response: - Network accessline 76
request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 77
with urllib.request.urlopen(request, timeout=30) as response: - Access to secretsline 123
[Credential-related source omitted] - Access to secretsline 124
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_02/SKILL.md
- Access to secretsline 25
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_02/scripts/fetch_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 29
[Credential-related source omitted] - Network accessline 31
[Credential-related source omitted] - Access to secretsline 31
[Credential-related source omitted] - Network accessline 32
[Credential-related source omitted] - Access to secretsline 32
[Credential-related source omitted] - Network accessline 33
req = urllib.request.Request(url, headers=headers, method="GET") - Network accessline 34
try: with urllib.request.urlopen(req, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 35
with urllib.request.urlopen(req, timeout=20) as response: - Network accessline 73
def main() -> int: parser = argparse.ArgumentParser(description="Fetch allowed Investigation Review Hub endpoints.") parser.add_argument("base_url", help="Hub base URL, for example http://task-env:9017/") - Network accessline 74
parser = argparse.ArgumentParser(description="Fetch allowed Investigation Review Hub endpoints.") parser.add_argument("base_url", help="Hub base URL, for example http://task-env:9017/") parser.add_argument("--matter-id", help="Matte - Network accessline 75
parser.add_argument("base_url", help="Hub base URL, for example http://task-env:9017/") - Access to secretsline 89
[Credential-related source omitted] - Access to secretsline 91
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/naive/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_01/scripts/fetch_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 59
data = json.dumps(body).encode("utf-8") request_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=request_headers, method=method) - Network accessline 60
request_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=request_headers, method=method) with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 61
req = urllib.request.Request(url, data=data, headers=request_headers, method=method) - Network accessline 62
with urllib.request.urlopen(req, timeout=20) as resp: - Access to secretsline 107
[Credential-related source omitted] - Access to secretsline 111
[Credential-related source omitted] - Access to secretsline 112
[Credential-related source omitted] - Access to secretsline 148
[Credential-related source omitted] - Access to secretsline 156
[Credential-related source omitted] - Access to secretsline 171
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_01/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_02/scripts/hub_extract.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
def parse_args() -> argparse.Namespace: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Hub base URL, for example http://task-env:9017/") - Network accessline 37
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", required=True, help="Hub base URL, for example http://task-env:9017/") parser.add_argument("--matter-id", required=True, help="Matter ID to extr - Network accessline 38
parser.add_argument("--base-url", required=True, help="Hub base URL, for example http://task-env:9017/") - Access to secretsline 40
[Credential-related source omitted] - Access to secretsline 46
[Credential-related source omitted] - Access to secretsline 52
[Credential-related source omitted] - Access to secretsline 53
[Credential-related source omitted] - Network accessline 55
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 56
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 57
with urllib.request.urlopen(request, timeout=30) as response: - Access to secretsline 103
[Credential-related source omitted] - Access to secretsline 105
[Credential-related source omitted] - Access to secretsline 133
[Credential-related source omitted] - Access to secretsline 135
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_03/scripts/fetch_hub_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 36
[Credential-related source omitted] - Network accessline 42
[Credential-related source omitted] - Access to secretsline 42
[Credential-related source omitted] - Network accessline 43
[Credential-related source omitted] - Access to secretsline 43
[Credential-related source omitted] - Network accessline 44
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 45
try: with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode("utf-8")) - Network accessline 46
with urllib.request.urlopen(req, timeout=30) as resp: - Access to secretsline 85
[Credential-related source omitted] - Access to secretsline 90
[Credential-related source omitted] - Access to secretsline 104
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_017/fewshot/opencode/fewshot_attempt_03/scripts/validate_structured_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_02/scripts/payment_schedule.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/cc/fewshot_attempt_03/scripts/collect_portal_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_01/scripts/payment_schedule.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_01/scripts/portal_collect.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
def fetch(base_url, endpoint_key, params=None): - Network accessline 40
if query: url = f"{url}?{query}" request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 41
url = f"{url}?{query}" request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 42
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 43
try: with urllib.request.urlopen(request, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 44
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 67
for case_number in unique(args.case): params = {"case_number": case_number} snapshot["cases"][case_number] = fetch(args.base_url, "cases", params) - Network accessline 68
params = {"case_number": case_number} snapshot["cases"][case_number] = fetch(args.base_url, "cases", params) snapshot["charges"][case_number] = fetch(args.base_url, "charges", params) - Network accessline 69
snapshot["cases"][case_number] = fetch(args.base_url, "cases", params) - Network accessline 70
snapshot["charges"][case_number] = fetch(args.base_url, "charges", params) - Network accessline 71
snapshot["docket_entries"][case_number] = fetch(args.base_url, "docket_entries", params) - Network accessline 73
for citation_number in unique(args.citation): snapshot["citations"][citation_number] = fetch( - Network accessline 74
snapshot["citations"][citation_number] = fetch( - Network accessline 80
for petition_id in unique(args.petition): snapshot["financial_petitions"][petition_id] = fetch( - Network accessline 81
snapshot["financial_petitions"][petition_id] = fetch( - Network accessline 87
for jurisdiction_code in unique(args.jurisdiction): params = {"jurisdiction_code": jurisdiction_code} snapshot["jurisdictions"][jurisdiction_code] = fetch(args.base_url, "jurisdictions", params) - Network accessline 88
params = {"jurisdiction_code": jurisdiction_code} snapshot["jurisdictions"][jurisdiction_code] = fetch(args.base_url, "jurisdictions", params) snapshot["fee_schedules"][jurisdiction_code] = fetch(args.base_url, "fee_schedule - Network accessline 89
snapshot["jurisdictions"][jurisdiction_code] = fetch(args.base_url, "jurisdictions", params) - Network accessline 90
snapshot["fee_schedules"][jurisdiction_code] = fetch(args.base_url, "fee_schedules", params) - Network accessline 91
snapshot["payment_policies"][jurisdiction_code] = fetch(args.base_url, "payment_policies", params) - Network accessline 92
snapshot["forms"][jurisdiction_code] = fetch(args.base_url, "forms", params) - Network accessline 94
for query in unique(args.search): snapshot["search"][query] = fetch(args.base_url, "search", {"q": query}) - Network accessline 95
snapshot["search"][query] = fetch(args.base_url, "search", {"q": query})
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_02/scripts/installment_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/codex/fewshot_attempt_03/scripts/court_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_01/scripts/payment_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_01/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
def fetch(base_url: str, endpoint: str, params: dict[str, str]) -> dict: - Network accessline 35
def add_request(out: dict, label: str, base_url: str, endpoint: str, params: dict[str, str]) -> None: out.setdefault(label, []).append(fetch(base_url, endpoint, params)) - Network accessline 36
out.setdefault(label, []).append(fetch(base_url, endpoint, params))
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_02/scripts/court_packet_helpers.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/deepagents/fewshot_attempt_03/scripts/payment_schedule.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_01/scripts/court_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_02/scripts/portal_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
if params: url = f"{url}?{urllib.parse.urlencode(params)}" with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 42
url = f"{url}?{urllib.parse.urlencode(params)}" with urllib.request.urlopen(url, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 43
with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 140
def main() -> int: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Portal base URL, for example http://task-env:9018/") - Network accessline 141
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("base_url", help="Portal base URL, for example http://task-env:9018/") parser.add_argument("--case", dest="cases", action="append", default=[], help="Target c - Network accessline 142
parser.add_argument("base_url", help="Portal base URL, for example http://task-env:9018/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_018/fewshot/opencode/fewshot_attempt_03/scripts/court_packet_math.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 126
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 127
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 128
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_01/scripts/fetch_licensing_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 22
sep = "&" if "?" in url else "?" url = f"{url}{sep}limit={limit}" with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 23
url = f"{url}{sep}limit={limit}" with urllib.request.urlopen(url, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 24
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_02/scripts/fetch_casebook.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 18
if params: url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 19
url += "?" + urllib.parse.urlencode(params) req = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(req, timeout=20) as response: - Network accessline 20
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 21
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/cc/fewshot_attempt_02/scripts/validate_answer_template.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_01/SKILL.md
- Network accessline 15
python3 skill/scripts/fetch_records.py --base-url http://task-env:9019 --domain contractor --ids C-001 C-002 - Network accessline 16
python3 skill/scripts/fetch_records.py --base-url http://task-env:9019 --domain liquor --ids L-001 --location-ids LOC-001 - Network accessline 17
python3 skill/scripts/fetch_records.py --base-url http://task-env:9019 --domain renewal --ids AL-001 AL-002
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_01/scripts/fetch_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 47
def fetch_json(base_url: str, endpoint: str) -> Any: url = base_url.rstrip("/") + endpoint req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 48
url = base_url.rstrip("/") + endpoint req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 49
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 50
try: with urllib.request.urlopen(req, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 51
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_02/scripts/collect_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 35
def get_json(base_url: str, path: str): url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 37
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 38
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/codex/fewshot_attempt_03/scripts/fetch_endpoint.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 35
def fetch_json(url: str) -> object: request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 36
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 37
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_01/scripts/fetch_licensing_records.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_02/scripts/fetch_evidence.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_records.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 21
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: - Network accessline 22
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 23
with urllib.request.urlopen(url, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_02/scripts/check_answer_template.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_02/scripts/fetch_licensing_data.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
[Credential-related source omitted] - Network accessline 31
headers["X-Task-Token"] = token req = urllib.request.Request(url, data=data, method=method, headers=headers) with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 32
req = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 33
with urllib.request.urlopen(req, timeout=30) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/naive/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_02/scripts/fetch_records.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_019/fewshot/opencode/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_01/SKILL.md
- Network accessline 42
curl -sS -X POST "$TASK_ENV_BASE_URL/api/query" \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_01/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 42
headers["content-type"] = "application/json" method = "POST" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 43
method = "POST" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 44
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 45
try: with urllib.request.urlopen(req, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 46
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_02/scripts/check_json_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_02/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 72
[Credential-related source omitted] - Network accessline 73
[Credential-related source omitted] - Access to secretsline 73
[Credential-related source omitted] - Network accessline 74
or read_base_from_environment_access() or "http://task-env:9020/" ) - Network accessline 75
or "http://task-env:9020/"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/cc/fewshot_attempt_03/scripts/collect_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 36
def fetch_json(base_url: str, path: str) -> dict[str, Any]: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 37
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 38
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 39
try: with urllib.request.urlopen(request, timeout=20) as response: raw = response.read().decode("utf-8") - Network accessline 40
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_01/scripts/fetch_deal_context.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 34
def get_json(base_url, path, timeout): url = build_url(base_url, path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 35
url = build_url(base_url, path) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 36
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 37
try: with urllib.request.urlopen(request, timeout=timeout) as response: body = response.read().decode("utf-8") - Network accessline 38
with urllib.request.urlopen(request, timeout=timeout) as response: - Access to secretsline 58
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_02/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/codex/fewshot_attempt_03/SKILL.md
- Network accessline 31
BASE="$TASK_ENV_BASE_URL" DEAL_ID="the_deal_id_from_the_prompt" curl -sS "$BASE/api/deals/$DEAL_ID" > /tmp/deal.json - Network accessline 32
DEAL_ID="the_deal_id_from_the_prompt" curl -sS "$BASE/api/deals/$DEAL_ID" > /tmp/deal.json curl -sS "$BASE/api/deals/$DEAL_ID/terms" > /tmp/terms.json - Network accessline 33
curl -sS "$BASE/api/deals/$DEAL_ID" > /tmp/deal.json - Network accessline 34
curl -sS "$BASE/api/deals/$DEAL_ID/terms" > /tmp/terms.json
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_01/scripts/workbench_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 42
req = urllib.request.Request(url, data=data, method=method, headers=headers) - Network accessline 43
try: with urllib.request.urlopen(req, timeout=20) as resp: raw = resp.read().decode("utf-8") - Network accessline 44
with urllib.request.urlopen(req, timeout=20) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_02/scripts/check_json_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_02/scripts/collect_workbench_records.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_deal_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
def _get_json(base_url: str, path: str) -> tuple[int, Any]: url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 31
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 32
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 33
try: with urllib.request.urlopen(request, timeout=20) as response: body = response.read().decode("utf-8") - Network accessline 34
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_01/scripts/check_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_01/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
def get_json(base_url: str, path: str): url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 31
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 32
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 33
try: with urllib.request.urlopen(request, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 34
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 45
[Credential-related source omitted] - Network accessline 46
[Credential-related source omitted] - Network accessline 47
request = urllib.request.Request( - Network accessline 52
) try: with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 53
try: with urllib.request.urlopen(request, timeout=20) as response: data = response.read().decode("utf-8") - Network accessline 54
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_02/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 47
body = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, method=method, headers=headers) - Network accessline 48
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, method=method, headers=headers) try: - Network accessline 49
req = urllib.request.Request(url, data=body, method=method, headers=headers) - Network accessline 50
try: with urllib.request.urlopen(req, timeout=20) as resp: raw = resp.read().decode("utf-8") - Network accessline 51
with urllib.request.urlopen(req, timeout=20) as resp: - Network accessline 102
def main() -> int: parser = argparse.ArgumentParser(description="Fetch M&A deal-workbench evidence as JSON.") parser.add_argument("--base-url", required=True, help="Workbench base URL, such as http://task-env:9020/") - Network accessline 103
parser = argparse.ArgumentParser(description="Fetch M&A deal-workbench evidence as JSON.") parser.add_argument("--base-url", required=True, help="Workbench base URL, such as http://task-env:9020/") parser.add_argument("--deal-id", r - Network accessline 104
parser.add_argument("--base-url", required=True, help="Workbench base URL, such as http://task-env:9020/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_03/SKILL.md
- Network accessline 12
1. Read the user prompt and `input/payloads/answer_template.json` before fetching data. 2. Extract the `deal_id`, client side, transaction type, named playbook or policy, requested categories, required ordering, units, precision, and enum s - Network accessline 13
2. Extract the `deal_id`, client side, transaction type, named playbook or policy, requested categories, required ordering, units, precision, and enum spellings. 3. Resolve `<TASK_ENV_BASE_URL>` from the prompt or environment. If it is not - Network accessline 14
3. Resolve `<TASK_ENV_BASE_URL>` from the prompt or environment. If it is not explicit, try `TASK_ENV_BASE_URL`, then `http://task-env:9020/`.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_020/fewshot/opencode/fewshot_attempt_03/scripts/fetch_workbench.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
def get_json(base_url: str, path: str) -> object: url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 38
url = urllib.parse.urljoin(base_url, path.lstrip("/")) request = urllib.request.Request(url, headers={"Accept": "application/json"}) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 39
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 89
parser.add_argument( "--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9020/"), - Network accessline 90
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9020/"), help="Task environment base URL", - Network accessline 91
[Credential-related source omitted] - Access to secretsline 91
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_01/scripts/hub_client.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 58
data = json.dumps(payload).encode("utf-8") req_headers.setdefault("Content-Type", "application/json") request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 59
req_headers.setdefault("Content-Type", "application/json") request = urllib.request.Request(url, data=data, headers=req_headers, method=method) try: - Network accessline 60
request = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 61
try: with urllib.request.urlopen(request) as response: body = response.read().decode("utf-8") - Network accessline 62
with urllib.request.urlopen(request) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_02/scripts/asteria_query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 50
body = json.dumps(payload).encode("utf-8") req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, method=method, headers=req_headers) - Network accessline 51
req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=body, method=method, headers=req_headers) try: - Network accessline 52
req = urllib.request.Request(url, data=body, method=method, headers=req_headers) - Network accessline 53
try: with urllib.request.urlopen(req, timeout=60) as resp: data = resp.read().decode("utf-8") - Network accessline 54
with urllib.request.urlopen(req, timeout=60) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/cc/fewshot_attempt_03/scripts/fetch_asteria.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 40
query = "?" + urllib.parse.urlencode(clean) url = base_url + path + query with urllib.request.urlopen(url) as response: - Network accessline 41
url = base_url + path + query with urllib.request.urlopen(url) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 42
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_01/scripts/contract_lint.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_01/scripts/hub_dump.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 40
[Credential-related source omitted] - Access to secretsline 41
[Credential-related source omitted] - Access to secretsline 42
[Credential-related source omitted] - Access to secretsline 47
[Credential-related source omitted] - Access to secretsline 48
[Credential-related source omitted] - Network accessline 73
data = json.dumps(body).encode("utf-8") req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 74
req_headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=req_headers, method=method) try: - Network accessline 75
req = urllib.request.Request(url, data=data, headers=req_headers, method=method) - Network accessline 76
try: with urllib.request.urlopen(req, timeout=60) as response: payload = response.read().decode("utf-8") - Network accessline 77
with urllib.request.urlopen(req, timeout=60) as response: - Access to secretsline 126
[Credential-related source omitted] - Access to secretsline 166
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_02/scripts/fetch_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
query["offset"] = str(offset) url = f"{base_url}{endpoint}?{urllib.parse.urlencode(query)}" with urllib.request.urlopen(url) as response: - Network accessline 38
url = f"{base_url}{endpoint}?{urllib.parse.urlencode(query)}" with urllib.request.urlopen(url) as response: payload = json.load(response) - Network accessline 39
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/codex/fewshot_attempt_03/scripts/fetch_hub.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 46
headers["Authorization"] = f"Bearer {credentials}" headers["X-API-Key"] = credentials request = urllib.request.Request(url, headers=headers) - Network accessline 47
headers["X-API-Key"] = credentials request = urllib.request.Request(url, headers=headers) try: - Network accessline 48
request = urllib.request.Request(url, headers=headers) - Network accessline 49
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 50
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_01/scripts/fetch_hub_bundle.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 37
query = "?" + urllib.parse.urlencode(params) url = base_url.rstrip("/") + path + query req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 38
url = base_url.rstrip("/") + path + query req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 39
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 40
try: with urllib.request.urlopen(req) as resp: return json.load(resp) - Network accessline 41
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_02/scripts/asteria_audit.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 107
if params: url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: - Network accessline 108
url += "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: data = json.load(response) - Network accessline 109
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 24
Use `--header 'Name: value'` only when `environment_access.md` supplies credentials. If using curl manually, pass `collection=<collection_id>` to data and source-snapshot endpoints and page with `limit`/`offset` until `offset + limit >= tot
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/deepagents/fewshot_attempt_03/scripts/fetch_hub_collection.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 47
if query: url = f"{url}?{query}" req = urllib.request.Request(url, headers=headers) - Network accessline 48
url = f"{url}?{query}" req = urllib.request.Request(url, headers=headers) try: - Network accessline 49
req = urllib.request.Request(url, headers=headers) - Network accessline 50
try: with urllib.request.urlopen(req, timeout=30) as response: body = response.read().decode("utf-8") - Network accessline 51
with urllib.request.urlopen(req, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/naive/fewshot_attempt_03/scripts/asteria_export.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_01/scripts/asteria_hub_audit.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 89
query["offset"] = offset url = f"{self.base_url}{endpoint}?{urllib.parse.urlencode(query)}" with urllib.request.urlopen(url) as resp: - Network accessline 90
url = f"{self.base_url}{endpoint}?{urllib.parse.urlencode(query)}" with urllib.request.urlopen(url) as resp: payload = json.load(resp) - Network accessline 91
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_02/SKILL.md
- Access to secretsline 38
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_02/scripts/check_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_02/scripts/hub_query.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 43
[Credential-related source omitted] - Access to secretsline 50
[Credential-related source omitted] - Access to secretsline 51
[Credential-related source omitted] - Access to secretsline 52
[Credential-related source omitted] - Network accessline 89
merged_headers = dict(headers) merged_headers.update(auth) req = urllib.request.Request(url, data=payload, headers=merged_headers, method=method) - Network accessline 90
merged_headers.update(auth) req = urllib.request.Request(url, data=payload, headers=merged_headers, method=method) try: - Network accessline 91
req = urllib.request.Request(url, data=payload, headers=merged_headers, method=method) - Network accessline 92
try: with urllib.request.urlopen(req, timeout=60) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 93
with urllib.request.urlopen(req, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_021/fewshot/opencode/fewshot_attempt_03/scripts/asteria_collect.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 74
query = urllib.parse.urlencode({k: v for k, v in params.items() if v is not None}) url = f"{url}?{query}" req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 75
url = f"{url}?{query}" req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 76
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 77
try: with urllib.request.urlopen(req, timeout=30) as resp: body = resp.read().decode("utf-8") - Network accessline 78
with urllib.request.urlopen(req, timeout=30) as resp: - Network accessline 126
def main(argv: list[str]) -> int: parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", help="Hub base URL, for example http://host:port/") - Network accessline 127
parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--base-url", help="Hub base URL, for example http://host:port/") parser.add_argument("--env-file", type=Path, help="environment_access.md path") - Network accessline 128
parser.add_argument("--base-url", help="Hub base URL, for example http://host:port/")
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/SKILL.md
- Access to secretsline 13
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 36
[Credential-related source omitted] - Access to secretsline 37
[Credential-related source omitted] - Access to secretsline 46
[Credential-related source omitted] - Access to secretsline 47
[Credential-related source omitted] - Network accessline 60
data = json.dumps(payload).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 61
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 62
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 63
try: with urllib.request.urlopen(req) as resp: raw = resp.read().decode("utf-8") - Network accessline 64
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_01/scripts/check_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_02/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 17
def base_url(value: str | None = None) -> str: url = value or os.environ.get("TASK_ENV_BASE_URL") or "http://task-env:9022/" - Network accessline 18
[Credential-related source omitted] - Access to secretsline 18
[Credential-related source omitted] - Access to secretsline 23
[Credential-related source omitted] - Network accessline 30
def request_json(method: str, path: str, body: Any = None, *, root: str | None = None) -> Any: data = None if body is None else json.dumps(body).encode("utf-8") req = urllib.request.Request( - Network accessline 31
data = None if body is None else json.dumps(body).encode("utf-8") req = urllib.request.Request( f"{base_url(root)}{path}", - Network accessline 32
req = urllib.request.Request( - Network accessline 37
) try: with urllib.request.urlopen(req, timeout=60) as response: - Network accessline 38
try: with urllib.request.urlopen(req, timeout=60) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 39
with urllib.request.urlopen(req, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_03/references/api-and-schema.md
- Network accessline 6
[Credential-related source omitted] - Access to secretsline 6
[Credential-related source omitted] - Network accessline 7
[Credential-related source omitted] - Access to secretsline 7
[Credential-related source omitted] - Network accessline 13
[Credential-related source omitted] - Access to secretsline 13
[Credential-related source omitted] - Access to secretsline 14
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/cc/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_01/references/api.md
- Access to secretsline 15
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 15
DEFAULT_BASE_URL = "http://task-env:9022/" - Access to secretsline 19
[Credential-related source omitted] - Network accessline 44
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 45
try: with urllib.request.urlopen(request) as response: raw = response.read().decode("utf-8") - Network accessline 46
with urllib.request.urlopen(request) as response: - Access to secretsline 84
[Credential-related source omitted] - Access to secretsline 104
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_02/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 12
def base_url(value): raw = value or os.environ.get("TASK_ENV_BASE_URL") or "http://task-env:9022/" - Network accessline 13
[Credential-related source omitted] - Access to secretsline 13
[Credential-related source omitted] - Access to secretsline 27
[Credential-related source omitted] - Network accessline 32
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 33
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 34
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 35
try: with urllib.request.urlopen(req) as resp: payload = resp.read().decode("utf-8") - Network accessline 36
with urllib.request.urlopen(req) as resp: - Access to secretsline 55
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_03/references/atlas-patterns.md
- Access to secretsline 7
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_03/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
[Credential-related source omitted] - Access to secretsline 41
[Credential-related source omitted] - Network accessline 42
or env_access.get("base_url") or "http://task-env:9022/" ) - Network accessline 43
or "http://task-env:9022/" - Access to secretsline 45
[Credential-related source omitted] - Access to secretsline 46
[Credential-related source omitted] - Network accessline 59
request = urllib.request.Request( - Network accessline 64
) try: with urllib.request.urlopen(request, timeout=60) as response: - Network accessline 65
try: with urllib.request.urlopen(request, timeout=60) as response: payload = response.read().decode("utf-8") - Network accessline 66
with urllib.request.urlopen(request, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/codex/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_01/SKILL.md
- Access to secretsline 27
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 14
def _base_url() -> str: base = os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9022/") - Network accessline 15
[Credential-related source omitted] - Access to secretsline 15
[Credential-related source omitted] - Access to secretsline 31
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 17
DEFAULT_BASE_URL = "http://task-env:9022/" - Access to secretsline 30
[Credential-related source omitted] - Access to secretsline 43
[Credential-related source omitted] - Network accessline 51
request = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 52
try: with urllib.request.urlopen(request) as response: body = response.read().decode("utf-8") - Network accessline 53
with urllib.request.urlopen(request) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_03/SKILL.md
- Network accessline 28
- `TASK_ENV_BASE_URL`, defaulting to `http://task-env:9022`
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/deepagents/fewshot_attempt_03/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 14
def base_url() -> str: return os.environ.get("TASK_ENV_BASE_URL", "http://task-env:9022").rstrip("/") - Network accessline 15
[Credential-related source omitted] - Access to secretsline 15
[Credential-related source omitted] - Access to secretsline 19
[Credential-related source omitted] - Network accessline 32
req = urllib.request.Request(base_url() + path, data=data, headers=headers) - Network accessline 33
try: with urllib.request.urlopen(req, timeout=60) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 34
with urllib.request.urlopen(req, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 43
[Credential-related source omitted] - Access to secretsline 44
[Credential-related source omitted] - Network accessline 60
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(base_url + path, data=data, headers=headers, method=method) - Network accessline 61
headers["Content-Type"] = "application/json" req = urllib.request.Request(base_url + path, data=data, headers=headers, method=method) try: - Network accessline 62
req = urllib.request.Request(base_url + path, data=data, headers=headers, method=method) - Network accessline 63
try: with urllib.request.urlopen(req) as response: raw = response.read().decode("utf-8") - Network accessline 64
with urllib.request.urlopen(req) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_02/SKILL.md
- Network accessline 19
curl -sS \
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_03/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/naive/fewshot_attempt_03/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 41
headers["Content-Type"] = "application/json" body = json.dumps(payload).encode("utf-8") request = urllib.request.Request(url, data=body, method=method, headers=headers) - Network accessline 42
body = json.dumps(payload).encode("utf-8") request = urllib.request.Request(url, data=body, method=method, headers=headers) try: - Network accessline 43
request = urllib.request.Request(url, data=body, method=method, headers=headers) - Network accessline 44
try: with urllib.request.urlopen(request) as response: raw = response.read().decode("utf-8") - Network accessline 45
with urllib.request.urlopen(request) as response: - Access to secretsline 58
[Credential-related source omitted] - Access to secretsline 59
[Credential-related source omitted] - Access to secretsline 60
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01/SKILL.md
- Network accessline 33
The helper reads `TASK_ENV_API_TOKEN` and uses `TASK_ENV_BASE_URL` if set, otherwise `http://task-env:9022`.
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_01/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 13
DEFAULT_BASE_URL = "http://task-env:9022" - Access to secretsline 17
[Credential-related source omitted] - Access to secretsline 21
[Credential-related source omitted] - Network accessline 35
data = json.dumps(body).encode("utf-8") headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 36
headers["Content-Type"] = "application/json" req = urllib.request.Request(url, data=data, headers=headers, method=method) try: - Network accessline 37
req = urllib.request.Request(url, data=data, headers=headers, method=method) - Network accessline 38
try: with urllib.request.urlopen(req, timeout=60) as resp: raw = resp.read().decode("utf-8") - Network accessline 39
with urllib.request.urlopen(req, timeout=60) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02/SKILL.md
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 15
DEFAULT_BASE_URL = "http://task-env:9022/" - Access to secretsline 19
[Credential-related source omitted] - Access to secretsline 26
[Credential-related source omitted] - Network accessline 36
req = urllib.request.Request(base_url + path, data=data, headers=headers, method=method) - Network accessline 37
try: with urllib.request.urlopen(req) as resp: raw = resp.read().decode("utf-8") - Network accessline 38
with urllib.request.urlopen(req) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_02/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_03/scripts/atlas_api.py
- Executable scriptline 1
#!/usr/bin/env python3 - Access to secretsline 13
[Credential-related source omitted] - Access to secretsline 21
[Credential-related source omitted] - Network accessline 32
data = json.dumps(body).encode("utf-8") req_headers["Content-Type"] = "application/json" request = urllib.request.Request(f"{base_url}{endpoint}", data=data, headers=req_headers, method=method) - Network accessline 33
req_headers["Content-Type"] = "application/json" request = urllib.request.Request(f"{base_url}{endpoint}", data=data, headers=req_headers, method=method) try: - Network accessline 34
request = urllib.request.Request(f"{base_url}{endpoint}", data=data, headers=req_headers, method=method) - Network accessline 35
try: with urllib.request.urlopen(request) as response: raw = response.read() - Network accessline 36
with urllib.request.urlopen(request) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_022/fewshot/opencode/fewshot_attempt_03/scripts/validate_answer.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01/SKILL.md
- Network accessline 53
python - <<'PY' from scripts.pho_math import download_portal_csvs base_url = "http://task-env:9023" - Network accessline 54
from scripts.pho_math import download_portal_csvs base_url = "http://task-env:9023" tables = download_portal_csvs(base_url, ["states", "state_health"]) - Network accessline 55
base_url = "http://task-env:9023"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_01/scripts/pho_math.py
- Executable scriptline 1
Executable file - Network accessline 28
query = urllib.parse.urlencode({"dataset": dataset, "format": "csv"}) url = f"{root}/download?{query}" with urllib.request.urlopen(url) as resp: - Network accessline 29
url = f"{root}/download?{query}" with urllib.request.urlopen(url) as resp: text = resp.read().decode("utf-8") - Network accessline 30
with urllib.request.urlopen(url) as resp:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_02/scripts/check_answer_shape.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_02/scripts/pho_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
query = urllib.parse.urlencode({"dataset": dataset, "format": "csv"}) url = urllib.parse.urljoin(base, "download") + "?" + query with urllib.request.urlopen(url, timeout=60) as response: - Network accessline 31
url = urllib.parse.urljoin(base, "download") + "?" + query with urllib.request.urlopen(url, timeout=60) as response: body = response.read().decode("utf-8") - Network accessline 32
with urllib.request.urlopen(url, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/cc/fewshot_attempt_03/scripts/pho_common.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 32
"download?" + urllib.parse.urlencode({"dataset": dataset, "format": "csv"}), ) with urllib.request.urlopen(url) as response: - Network accessline 33
) with urllib.request.urlopen(url) as response: text = response.read().decode("utf-8") - Network accessline 34
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_01/scripts/pho_math.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02/scripts/fetch_portal.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 42
url = urllib.parse.urljoin(normalize_base_url(base_url), f"download?{query}") out_path = out_dir / f"{dataset}.csv" with urllib.request.urlopen(url, timeout=60) as response: - Network accessline 43
out_path = out_dir / f"{dataset}.csv" with urllib.request.urlopen(url, timeout=60) as response: body = response.read() - Network accessline 44
with urllib.request.urlopen(url, timeout=60) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_02/scripts/pho_common.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/scripts/pho_audit_utils.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 51
url = url + "?" + urllib.parse.urlencode({"dataset": dataset, "format": "csv"}) out_path = os.path.join(out_dir, f"{dataset}.csv") with urllib.request.urlopen(url) as response, open(out_path, "wb") as f: - Network accessline 52
out_path = os.path.join(out_dir, f"{dataset}.csv") with urllib.request.urlopen(url) as response, open(out_path, "wb") as f: f.write(response.read()) - Network accessline 53
with urllib.request.urlopen(url) as response, open(out_path, "wb") as f:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_01/scripts/pho_audit_utils.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/deepagents/fewshot_attempt_02/scripts/pho_tools.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 29
params[k] = v url = base + "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: - Network accessline 30
url = base + "?" + urllib.parse.urlencode(params) with urllib.request.urlopen(url) as response: text = response.read().decode("utf-8") - Network accessline 31
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/naive/fewshot_attempt_01/scripts/pho_stats.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_01/references/protocol-profiles.md
- Dangerous shell operationline 129
magnitude = SE_b * sqrt(df * rY * rM / (1 - rM))
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_01/scripts/pho_common.py
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_02/SKILL.md
- Network accessline 29
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=state_health&format=csv" - Network accessline 30
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=state_socioeconomic&format=csv" - Network accessline 31
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=county_health&format=csv" - Network accessline 32
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=county_socioeconomic&format=csv" - Network accessline 33
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=country_indicators&format=csv" - Network accessline 34
curl -fsS "$TASK_ENV_BASE_URL/download?dataset=revisions&format=csv"
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_02/scripts/pho_audit_utils.py
- Executable scriptline 1
Executable file
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/opencode/fewshot_attempt_03/scripts/pho_audit_helpers.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 40
url = urllib.parse.urljoin(base_url.rstrip("/") + "/", "download") url = url + "?" + urllib.parse.urlencode(query, doseq=True) with urllib.request.urlopen(url) as response: - Network accessline 41
url = url + "?" + urllib.parse.urlencode(query, doseq=True) with urllib.request.urlopen(url) as response: text = response.read().decode("utf-8") - Network accessline 42
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_01/scripts/env_review.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 81
def get_json(base_url: str, path: str) -> dict: url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 82
url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 83
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 84
try: with urllib.request.urlopen(req, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 85
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/cc/fewshot_attempt_03/scripts/portfolio_env_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 105
query = "?" + urllib.parse.urlencode(params, doseq=True) url = f"{base_url}{path}{query}" with urllib.request.urlopen(url) as response: - Network accessline 106
url = f"{base_url}{path}{query}" with urllib.request.urlopen(url) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 107
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_01/scripts/env_audit.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 66
def fetch_json(base_url: str, path: str) -> dict: url = base_url.rstrip("/") + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 67
url = base_url.rstrip("/") + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 68
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 69
try: with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 70
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_02/scripts/fetch_env_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/codex/fewshot_attempt_03/scripts/portfolio_env_helper.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 52
url = base_url.rstrip("/") + path try: with urllib.request.urlopen(url) as response: - Network accessline 53
try: with urllib.request.urlopen(url) as response: return json.load(response) - Network accessline 54
with urllib.request.urlopen(url) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_01/scripts/portfolio_env.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 79
[Credential-related source omitted] - Network accessline 80
[Credential-related source omitted] - Network accessline 81
req = urllib.request.Request(url) - Network accessline 83
req.add_header("X-Env-Token", token) try: with urllib.request.urlopen(req, timeout=30) as response: - Network accessline 84
try: with urllib.request.urlopen(req, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 85
with urllib.request.urlopen(req, timeout=30) as response: - Access to secretsline 520
[Credential-related source omitted] - Access to secretsline 521
[Credential-related source omitted]
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_02/scripts/fetch_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 30
def fetch_json(base_url: str, path: str) -> Any: url = base_url.rstrip("/") + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 31
url = base_url.rstrip("/") + path request = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 32
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 33
try: with urllib.request.urlopen(request, timeout=20) as response: return json.load(response) - Network accessline 34
with urllib.request.urlopen(request, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/deepagents/fewshot_attempt_03/scripts/env_fetch.py
- Executable scriptline 1
#!/usr/bin/env python3
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/naive/fewshot_attempt_03/scripts/portfolio_tool.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 79
def fetch_json(base_url: str, path: str) -> Any: url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 80
url = base_url.rstrip("/") + path req = urllib.request.Request(url, headers={"Accept": "application/json"}) try: - Network accessline 81
req = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 82
try: with urllib.request.urlopen(req, timeout=20) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 83
with urllib.request.urlopen(req, timeout=20) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_02/scripts/fetch_environment.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 26
[Credential-related source omitted] - Network accessline 27
[Credential-related source omitted] - Network accessline 28
request = urllib.request.Request(url, headers={"Accept": "application/json"}) - Network accessline 29
[Credential-related source omitted] - Network accessline 30
request.add_header("X-Env-Token", token) with urllib.request.urlopen(request, timeout=30) as response: return json.loads(response.read().decode("utf-8")) - Network accessline 31
with urllib.request.urlopen(request, timeout=30) as response:
experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_024/fewshot/opencode/fewshot_attempt_03/scripts/fetch_env_snapshot.py
- Executable scriptline 1
#!/usr/bin/env python3 - Network accessline 26
[Credential-related source omitted] - Network accessline 27
headers["X-Env-Token"] = token request = urllib.request.Request(url, headers=headers) with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 28
request = urllib.request.Request(url, headers=headers) - Network accessline 29
with urllib.request.urlopen(request, timeout=20) as response: - Network accessline 38
"--base-url", default=os.environ.get("TASK_ENV_BASE_URL"), help="Task environment base URL, for example http://task-env:9024", - Network accessline 39
[Credential-related source omitted] - Access to secretsline 39
[Credential-related source omitted] - Network accessline 40
help="Task environment base URL, for example http://task-env:9024", - Access to secretsline 44
[Credential-related source omitted]
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
Declared and executed: 2 facts
- Contacts fonts.googleapis.com, which SKILL.md, the README and the plugin manifest do not mentionupstream/scripts/generate_report.py:39High confidence
- Contacts fonts.gstatic.com, which SKILL.md, the README and the plugin manifest do not mentionupstream/scripts/generate_report.py:40High confidence
Signal confidence: 498 facts
- Network access in prose of documentationupstream/LICENSE.txt:2Low confidence
- Network access in prose of documentationupstream/LICENSE.txt:4Low confidence
- Network access in prose of documentationupstream/LICENSE.txt:196Low confidence
- Executable script in code of a file that runsupstream/eval-viewer/generate_review.py:1High confidence
- Network access in code of a file that runsupstream/eval-viewer/generate_review.py:449High confidence
- Executable script in prose of a file that runsupstream/scripts/__init__.py:1Low confidence
- Executable script in code of a file that runsupstream/scripts/aggregate_benchmark.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/generate_report.py:1High confidence
- Network access in code of a file that runsupstream/scripts/generate_report.py:37High confidence
- Network access in code of a file that runsupstream/scripts/generate_report.py:38High confidence
- Network access in code of a file that runsupstream/scripts/generate_report.py:39High confidence
- Network access in code of a file that runsupstream/scripts/generate_report.py:40High confidence
- Network access in code of a file that runsupstream/scripts/generate_report.py:41High confidence
- Executable script in code of a file that runsupstream/scripts/improve_description.py:1High confidence
- Access to secrets in prose of a file that runsupstream/scripts/improve_description.py:6Low confidence
- Access to secrets in code of a file that runs: a variable, placeholder or call, not a stored valueupstream/scripts/improve_description.py:33Low confidence
- Executable script in code of a file that runsupstream/scripts/package_skill.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/quick_validate.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/run_eval.py:1High confidence
- Access to secrets in code of a file that runs: a variable, placeholder or call, not a stored valueupstream/scripts/run_eval.py:83Low confidence
- Executable script in code of a file that runsupstream/scripts/run_loop.py:1High confidence
- Executable script in prose of a file that runsupstream/scripts/utils.py:1Low confidence
- Network access in prose of documentationupstream/LICENSE.txt:2Low confidence
- Network access in prose of documentationupstream/LICENSE.txt:4Low confidence
- Network access in prose of documentationupstream/LICENSE.txt:196Low confidence
- Network access in code of documentationupstream/references/openai_yaml.md:20Low confidence
- Network access in code of documentationupstream/references/openai_yaml.md:22Low confidence
- Executable script in code of a file that runsupstream/scripts/generate_openai_yaml.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/init_skill.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/quick_validate.py:1High confidence
- Executable script in code of a file that runsupstream/scripts/init_skill.py:1High confidence
- Network access in prose of a file that runsupstream/scripts/init_skill.py:195Low confidence
- Executable script in code of a file that runsupstream/scripts/quick_validate.py:1High confidence
- Network access in prose of the SKILL.mdfewshot_attempt_01/SKILL.md:5Low confidence
- Network access in prose of the SKILL.mdreflect-3_attempt_01/SKILL.md:5Low confidence
- Network access in prose of the SKILL.mdself_attempt_01/SKILL.md:5Low confidence
- Network access in prose of the SKILL.mdself_attempt_02/SKILL.md:5Low confidence
- Network access in code of the SKILL.mdfewshot_attempt_03/SKILL.md:6High confidence
- Network access in prose of the SKILL.mdself_attempt_02/SKILL.md:5Low confidence
- Network access in prose of the SKILL.mdfewshot_attempt_02/SKILL.md:151Low confidence
- Network access in prose of the SKILL.mdfewshot_attempt_03/SKILL.md:5Low confidence
- Network access in prose of the SKILL.mdfewshot_attempt_03/SKILL.md:6Low confidence
- Network access in prose of the SKILL.mdself_attempt_01/SKILL.md:5Low confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:214High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:215High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:217High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:218High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:219High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:220High confidence
- Network access in code of the SKILL.mdself_attempt_02/SKILL.md:221High confidence
and 448 more
Only the first 500 facts are listed.
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md evaluation/eval_workspace/codex_skill_creator_comparison/creators/cc/upstream/SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.
