Menu

Individual skill

pho-observatory-audits

0upvotes

Solve Public Health Observatory Web portal audit tasks from analysis_request.json and answer_template.json, including state, county, and country release resolution; cohort construction; fixed-effect, GMM, ridge, elastic-net, bootstrap, conformal, PCA, clustering, sensitivity, source-perturbation, and controlled-decision modules. Use when a task provides a PHO/Public Health Observatory portal base URL plus registered audit request/template files and asks for exactly one JSON answer.

experiments/codex_skill_creator_comparison_gpt5_5_xhigh/reports/skills/task_group_023/fewshot/codex/fewshot_attempt_03/SKILL.md

Read pinned source

Source commit 5082a777379156a048881e94940d640702e5e09a

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
No
Repository scan complete
No
Skill scan complete
No
Observed behavior
Not checked yet
Last checked
Nov 6, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
121
Repository last updated
Oct 29, 2026
License
Apache-2.0

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# PHO Observatory Audits

## Core Workflow

1. Parse `analysis_request.json`, `answer_template.json`, and the user prompt. Treat the template as the output contract and the request as the binding analytical contract.
2. Resolve the effective request before touching data. Apply exact `protocol_id` profiles and request overrides only when they match exactly; arrays replace whole arrays, objects deep-merge by exact key, scalars replace exact paths, and unknown aliases are errors.
3. Read [references/portal.md](references/portal.md) before data access. Fetch evidence only from the task's PHO portal base URL, usually via `/download?dataset=<name>&format=csv`.
4. If the request has one of the known `protocol_id` values or the country burden template, read [references/protocol_profiles.md](references/protocol_profiles.md) and use the matching module recipe.
5. Use [references/numeric_methods.md](references/numeric_methods.md) for deterministic formulas and ordering rules. Use `scripts/pho_audit_utils.py` for portable CSV fetching, release selection, PRNGs, quantiles, linear-model covariance, PCA, clustering, and JSON rounding helpers.
6. Build a task-specific solver script in the task workspace. Keep all intermediate statistics unrounded; round only the final JSON fields to the precision declared by the template/request.
7. Validate the finished object mechanically: required top-level keys, required nested keys, list lengths, list ordering, numeric precision, enum values, booleans, integer fields, and no narrative outside JSON.

## Data Rules
Back to top