Experience distiller
Distill a terminal reviewed run into reusable, provenance-linked experience. Use when merge, rejection, or human closure is complete, all evidence exists, and no operational action remains.From its SKILL.md
npx -y skills add YZYY95K/test --skill experience-distillerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.5 KB, 264 tokens by cl100k_base, as published. Nobody here has run it
Experience Distiller
Create a safe ExperiencePattern; do not reopen or reinterpret the decision.
Procedure
- Require a terminal reviewed outcome, trace ID, artifact digests, and policy decision.
- Redact credentials, personal data, and unnecessary proprietary source.
- Summarize symptom, root cause, attempted strategy, result, validation, and reusable lesson. Preserve uncertainty and failed approaches.
- Link large artifacts by immutable reference instead of copying them.
- Deduplicate existing patterns and update provenance rather than amplifying duplicates.
- Run
python scripts/validate.py output <artifact.json>and store only after every evidence and redaction gate passes.
Decision rules
- Missing provenance is rejected.
- Redaction uncertainty is quarantined for HumanReviewer.
- Storage outage emits a retryable failure; it does not alter the completed run.
Boundaries
Do not store raw secrets, personal data, full private files, unreviewed runs, or unsupported causal claims. Do not trigger new coding, testing, or merge work.
Read the contract for storage and quarantine rules. Read examples to calibrate useful abstraction without source leakage.
What ships with it: 4 files
6.1 KB alongside SKILL.md, 1 of them executable
agents/
- openai.yaml221 B
references/
- contract.yaml3.0 KB
- examples.md578 B
scripts/
- validate.pyruns2.3 KB