skills
AGENTS.md0 starsChanged 33 days agoMITView on GitHub
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 31,774 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.
- 1on1Skill
- context-reviewSkill
- gmail-inboxSkill
- lamport-problem-statementSkill
- learning-labSkill
- logseq-db-querySkill
- nix-sandboxSkill
- obsidian-zettelkastenSkill
- peer-reviewSkill
- portkey-statusSkill
- proton-inboxSkill
- teams-extractorSkill
- theory-of-constraintsSkill
The file
We split this AGENTS.md into the rules it gives an agent
6 rules. About 0 are common, 3 are distinctive, and 3 are 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.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 2
Do not commit without explicit user approval.
- Very specific: line 8
Follow the agent skills spec (site, github)
- Distinctive: line 9
Follow Anthropic's skill authoring best practices
- Distinctive: line 10
Write skill instructions as agent directives (procedural, imperative) — not human documentation prose.
- Very specific: line 11
Use annotated YAML with ? for optional fields and inline comments for constraints as the schema format — no tooling, clean, precise.
- Very specific: line 13
No hardcoded personal paths (e.g.
≈ “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
Git
- Do not commit without explicit user approval.
Quality
- After any measurable change to a skill (new skill, updated instructions, added format files, refactored steps, changed triggers), run the
skill-creatorskill, if available, to evaluate and improve it. Use your judgment on what counts as measurable — when in doubt, ask the user whether to run it.
Skill Authoring Conventions
- Follow the agent skills spec (site, github)
- Follow Anthropic's skill authoring best practices
- Write skill instructions as agent directives (procedural, imperative) — not human documentation prose. Red flag: explanatory asides addressed to a reader rather than an executor.
- Use annotated YAML with
?for optional fields and inline comments for constraints as the schema format — no tooling, clean, precise. - Extra resource files belong under
resources/(or a subdir of it), not at the skill root. - No hardcoded personal paths (e.g.
/Users/<username>/...on macOS,/home/<username>/...on Linux) in skill files. Use variables, placeholders, or a single configurable comment at the top of the file instead.
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.