Menu

Individual skill

Intents

Not available

LifeOS/install/skills/HillClimb/Intents/Skill.md

Read pinned source

Source commit b7efec97d8c3885feee285bee468c35602fb0606

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 2, 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.

Standard SKILL.md format has not been attested at this commit. No installation command is available; inspect the source.

Repository context

Repository stars (not skill ratings)
19,765
Repository last updated
Nov 5, 2026
License
MIT

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

Declared tags are partial or unavailable.

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
# Skill

Climb one skill until its outputs stop drawing the complaints and misses that started the climb.

## Ideal State

- The skill's ideal state is named: its ISA if it has one, else its "done looks like" block, its standing checks (`Proof.ts` profile entries with `skill:<Name>`), and the asks it serves.
- A baseline exists from one or more of these: the skill's suite pass^k (`EvalRunner.ts`), its Errata complaint rate (`Errata.ts skills`), and its receipt verdicts. The climb records which one is the primary metric.
- The baseline came from a raw run file with a seeded train/test split, and `preflight` said the suite can see a win worth having. If it could not, the climb added cases or trials before round 1.
- A skill with no suite had one before the first experiment, built from its real misses through the Assay CreateSuite intent.
- The misses are real: Errata complaints joined to its outputs, false passes, failing suite cases, and transcripts where it went wrong. Each becomes a case through `FailureToCase.ts`.
- Candidate changes come from the varied proposers SKILL.md § Ideal State describes (`Intents/Ideate.md` produces them), fed only what that section allows.
- What was learned lands as a dated Gotcha, a new standing check, or a deterministic tool, never as a longer prompt (SKILL.md § Ideal State).
- The kept diff is committed, and Verify later reports whether the skill's complaint rate moved.