claude-code-kit
AGENTS.md2 starsChanged 13 days agoMITView on GitHub
For any agent (Claude Code, Codex, others) changing this kit:
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,843 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository13 other files it gives its agents.
- capability-builderSkill
- ci-reviewSkill
- daily-syncSkill
- handoffSkill
- insta360-montageSkill
- kbSkill
- pr-watchSkill
- remindersSkill
- reopen-sessionsSkill
- sprint-demoSkill
- tech-watchSkill
- test-evidenceSkill
- video-montageSkill
The file
We split this AGENTS.md into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 5
Read CONTRIBUTING.md first: how a part is laid out, described in parts.json, set up and tested.
- Very specific: line 9
Every part is a plugin of .claude-plugin/marketplace.json: bump the version of each plugin you change, or its users never get the change.
- Distinctive: line 10
No personal data in any file (names, emails, hosts, paths).
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.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
Working on this repo
For any agent (Claude Code, Codex, others) changing this kit:
- Read CONTRIBUTING.md first: how a part is laid out, described in
parts.json, set up and tested. - CONTRIBUTING.md holds the maintainer's rules: follow them and never edit that file; when a rule seems wrong or missing, say so to the person you work with.
- A mod follows its Rules for mods: folded by default, always grouped, the whole line folds and lights under the pointer, a ↗ back to the conversation, long text wraps, and it works on the oldest supported Claude Code build too.
- Everything must run on macOS (bash 3.2), Linux and Windows (Git Bash). Check before finishing:
bash tests/git-identity-guard.sh,bash tests/kit-update-check.sh,bash tests/install-sandbox.sh --autoandbash tests/plugins.sh(both must end on0 problem(s)),bash tests/plugin-versions.sh(every plugin you changed has a new version),claude plugin validate ., and for a modcd mods && claude plugin test <mod>. - Every part is a plugin of
.claude-plugin/marketplace.json: bump the version of each plugin you change, or its users never get the change. - No personal data in any file (names, emails, hosts, paths). Ask before committing and pushing.
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.