promote-repeated-session-knowledge-to-a-skill
Use when the same procedure is re-derived from scratch every session. Turns the repeatable steps into a skill with real commands and pitfalls.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 81
- Forks
- 17
01 前言 Front Matter
- name
- promote-repeated-session-knowledge-to-a-skill
- description
- Use when the same procedure is re-derived from scratch every session. Turns the repeatable steps into a skill with real commands and pitfalls.
02 资源文件 3
- 许可证LICENSE1.0 KB
- 参考文档SOURCE.md371 B
- 参考文档TRUST.auto.yaml2.2 KB
03 SKILL.md 正文
Promote Repeated Session Knowledge to a Skill
If you have worked out the same steps three times, the fourth time should be a lookup. Promotion turns a re-derived procedure into a stored one — but only when it is stable.
Procedure
- Confirm the repetition: find three separate sessions where the same steps were derived.
- Write the procedure down as the exact commands that worked, not a paraphrase.
- Include the environment: paths, versions, credential source, so it runs without re-discovery.
- Add the pitfalls you actually hit — a skill without its failure modes re-derives them.
- Give it a slug that states the action:
promote-repeated-session-knowledge-to-a-skill, notsession-tips. - Add a verification block: the command that proves the procedure worked.
- Test the skill on a clean context: run it as if you had never done it before.
- If a step needs judgement that varies by case, keep that as a branch, not a fixed command.
- Only promote procedures that have succeeded three times; a twice-used trick may still be wrong.
- Link it from the session notes so the next occurrence finds the skill instead of re-deriving.
Pitfalls
-
Promoting a procedure that worked once, freezing a lucky result into policy.
-
Writing the steps in prose so the next reader cannot run them verbatim.
-
Omitting the pitfalls, so the skill re-derives the same failure every time.
-
Encoding case-specific judgement as a fixed command, breaking it on the next input.
-
Promoting before the third occurrence, when the procedure is still changing each run.
-
Promoting the steps but not the environment, so the next run fails on a missing path.
-
Copying an example from another skill and leaving its commands, which do not apply.
-
Making the skill so general that it no longer states a runnable step.
Verification
ls skills/<slug>/SKILL.md && grep -c "^[0-9]\." skills/<slug>/SKILL.md
# passes when the skill exists and every step is a runnable command or an explicit branch
Report to the user: the slug created, the three sessions that showed the repetition, and the verification command.
04 同仓库更多 Skills
recipe-performance
Diagnose a specific MetaMask user flow's real-app performance and Sentry critical-user-flow telemetry with deterministic mm-harness recipes. Use for CUF boundaries, cold/warm/background-resume behavior, latency regressions, lifecycle tags, duplicate or blocking network requests, long tasks, layout/paint cost, screen redraws, React rerenders, or browser trace analysis where tests alone cannot prove the real app.
mexc-spot-websocket
MEXC Spot WebSocket workflows for public streams, protobuf handling, depth maintenance, listen keys, and private streams without exposing API keys.
mexc-spot-rest
MEXC Spot REST workflows for public data, signed account requests, orders, wallet/admin actions, and secure signing without exposing API keys.
mexc-futures-websocket
MEXC Futures WebSocket workflows for public streams, private account/order/position streams, ping/pong, subscriptions, and depth maintenance.
0xklen-version-and-refresh-sanctions-lists
暂无简介
verify-tests-cover-changed-lines
Use when a PR adds code that the existing suite may not exercise. Measures patch coverage — the share of newly changed lines hit by tests — before approving.