cookbook
CLAUDE.md0 starsChanged 3 months agoMITView on GitHub
Battle-tested coding lessons from real AI-agent sessions. FOSS, plain markdown, agent-agnostic. NOT a style guide or idiom reference — official docs own that. Every lesson here names the actual bug that taught it. If it doesn't have a scar, it doesn't belong. - Read AGENTS.md before touching any SKILL.md — contribution rules are binding. - Skills live as <topic>/SKILL.md with frontmatter (name, description). To use one elsewhere: copy into ~/.claude/skills/<name>/SKILL.md (global) or <project>/.claude/skills/<name>/SKILL.md (per-project). Auto-discovered. - LEARN-PROMPT.md = the update…
In this repository9 other files it gives its agents.
- AGENTS.mdAGENTS.md
- android-share-intentsSkill
- ci-verify-generated-codeSkill
- multi-agent-gitSkill
- anti-default-gatesSkill
- flutter-drift-androidSkill
- android-settings-wiringSkill
- rust-boundariesSkill
- windowsSkill
The file
We split this CLAUDE.md into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
NOT a style guide or idiom reference — official docs own that.
- Very specific: line 13
Read AGENTS.md before touching any SKILL.md — contribution rules are binding.
- Distinctive: line 23
No lesson without a concrete failure behind it.
- Distinctive: line 25
Update > duplicate; delete stale content in the same edit.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CLAUDE.md — cookbook
Battle-tested coding lessons from real AI-agent sessions. FOSS, plain markdown, agent-agnostic.
What this repo is
NOT a style guide or idiom reference — official docs own that. Every lesson here names the actual bug that taught it. If it doesn't have a scar, it doesn't belong.
For Claude Code sessions in this repo
- Read
AGENTS.mdbefore touching any SKILL.md — contribution rules are binding. - Skills live as
<topic>/SKILL.mdwith frontmatter (name,description). To use one elsewhere: copy into~/.claude/skills/<name>/SKILL.md(global) or<project>/.claude/skills/<name>/SKILL.md(per-project). Auto-discovered. LEARN-PROMPT.md= the update workflow. After analyzing any project session, follow it: dedupe → create-or-update → commitlearn(<topic>): <summary>.
Hard rules
- ≤100 lines per SKILL.md.
- No lesson without a concrete failure behind it.
- Nothing official docs already teach.
- Update > duplicate; delete stale content in the same edit.
- Polyglot topics →
cross/.
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.

