lesson-learned
Skill2 starsChanged 25 days agoMITView on GitHub
Use when a mistake, a print-qa FAIL, a CONTRADICTED fact-check row, a correction from the maker or an agent that missed its output format has just happened. Writes the post-mortem to lessons\YYYY-MM-DD-<project>-<slug>.md in the standard table and rebuilds the lessons index.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- compozy/compozytheir own version, first seen Nov 13, 2026
- imMamdouhaboammar/kaku-chatgpt-harnesstheir own version, first seen Oct 25, 2026
- sueryue/Claude-Skilllstheir own version, first seen Jul 28, 2026
- nbirkner/digital-braintheir own version, first seen Apr 12, 2026
- FelipeAngeli/skillstheir own version, first seen Apr 10, 2026
- pedronauck/skillstheir own version, first seen Apr 3, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- compozy/compozy★ 2.9ksame name
- pedronauck/skills★ 655same name
- FelipeAngeli/skills★ 2same name
- imMamdouhaboammar/kaku-chatgpt-harness★ 0same name
- nbirkner/digital-brain★ 0same name
- sueryue/Claude-Skillls★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository30 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- agent-reviewSkill
- ai-mesh-cleanSkill
- cad-buildSkill
- couponSkill
- curriculumSkill
- dfm-reviewSkill
- fileSkill
- gddSkill
- ip-checkSkill
- led-budgetSkill
- makerworld-publishSkill
- material-pickSkill
- mcp-new-toolSkill
- ortho-sheetSkill
- pi-power-budgetSkill
- preprint-checkSkill
- promo-planSkill
- qa-gateSkill
- render-criticSkill
- research-oneSkill
- research-topicSkill
- retopoSkill
- safety-sheetSkill
- sculptSkill
- slicer-questionSkill
- style-bibleSkill
- tech-packSkill
- unit-costSkill
- verifySkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 1 is distinctive, and 3 are very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 15
Write the file in the standard format (copy the layout of any existing lesson in lessons\):
- Very specific: line 27
Add a Sources: list with full URLs when the lesson corrects a fact.
- Distinctive: line 28
If the rule does not yet live in anything executable (hook, script check, agent rationalization row), write "not yet enforced" and name the …
- Very specific: line 29
Run .venv\Scripts\python.exe tools\build_indexes.py.
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Overview
Input: $ARGUMENTS, three fields separated by |: project slug, lesson slug, one-line description. Output: lessons\YYYY-MM-DD-<project>-<slug>.md (today's date) and a regenerated lessons\INDEX.md. Lead-run; writes nothing else.
When to use
In the same session as the failure: any print-qa FAIL, any fact-checker CONTRADICTED row, any agent returning without its required output, any correction from the maker, any instruction given in the wrong form. Not for filing facts (/file). If a lesson with the same root cause already exists, add a row there instead of a new file.
Steps
- Glob
lessons\*.mdand Grep the folder for the slug and the root cause. If an existing lesson covers it, append a row to its table and go to step 4. - Write the file in the standard format (copy the layout of any existing lesson in
lessons\):
Add a# Post-mortem: <title> (<YYYY-MM-DD>, <project or phase>) | What happened | Root cause | Rule now | |---|---|---| | <observable event, naming the file or print it touched, repo-relative path> | <why, one sentence that names the missing check> | <imperative rule with the number, condition, command or file that makes it checkable> | Fix applied: <what changed, paths>. Rule lives in: <path of CLAUDE.md, an agent or skill file, a tools\*.py or hook, or "not yet enforced">.Sources:list with full URLs when the lesson corrects a fact. - If the rule does not yet live in anything executable (hook, script check, agent rationalization row), write "not yet enforced" and name the follow-up. Never claim enforcement that does not exist.
- Run
.venv\Scripts\python.exe tools\build_indexes.py; confirm the new line inlessons\INDEX.md. - Reply with the lesson path, the rule in one line, and where it lives (or "not yet enforced" plus the follow-up).
Common rationalizations
| Rationalization | Reality |
| "It was a small slip, no lesson needed." | Small slips (one wrong drying time, one wrong unit) repeat. Every correction gets a row. |
| "The rule is 'be more careful'." | Not checkable. A rule names a number, a condition, a command or a file. |
| "I'll write the lesson at the end of the session." | Sessions end abruptly. Same moment, same session. |
| "Rule lives in: CLAUDE.md" when it does not yet. | Write "not yet enforced" and name the hook or script that would. |
Red flags
- A "Rule now" cell without a checkable element.
- A lesson that restates the event without a root cause.
build_indexes.pynot run.
Verification
- File exists at
lessons\YYYY-MM-DD-<project>-<slug>.mdwith the three-column table. - "Rule lives in:" names a path or says "not yet enforced" with a follow-up.
build_indexes.pyprinted the new lesson count;lessons\INDEX.mdlists the file.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

