agent-skills
AGENTS.md4 starsChanged 14 days agoMITView on GitHub
Public collection of installable AI agent skills for Claude Code and compatible agents. Each skill is a self-contained folder in skills/ with a SKILL.md at its root. Use branch main. Release notes belong only in the gh release create body; there is no CHANGELOG.md. Versions live only in git tags and .claude-plugin/marketplace.json metadata.version. There is no root package.json or pnpm version step.
- Commits and pushes
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 repository20 other files it gives its agents.
- agent-creatorSkill
- apexSkill
- audio-loopSkill
- award-designSkill
- brand-voiceSkill
- claude-mdSkill
- code-ultrareviewSkill
- design-systemSkill
- download-mediaSkill
- forgeSkill
- frontend-devSkill
- humanize-enSkill
- markitdownSkill
- notionSkill
- oneshotSkill
- scaffoldSkill
- suno-produceSkill
- ultrapexSkill
- video-loopSkill
- write-clear-readmeSkill
The file
We split this AGENTS.md into the rules it gives an agent
6 rules. About 0 are common, 1 is distinctive, and 5 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.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Very specific: line 12
Authoring or evaluating a skill: read .agents/rules/skill-authoring.md and use the official Anthropic skill-creator.
- Very specific: line 13
Frontmatter, layout, or portability: read .agents/rules/agentskills-spec.md.
- Distinctive: line 15
Adding, changing, or removing tests: read Testing for regression value, coverage preservation, and execution limits.
- Very specific: line 20
Run python3 -m unittest discover tests/ before reporting done and before commit.
- Very specific: line 21
After editing a canonical skill-*-rules.md block, run python3 scripts/sync_writing_rules.py and review the propagated diff.
- Very specific: line 25
Use branch main. Release notes belong only in the gh release create body.
≈ “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
Agent Skills
Public collection of installable AI agent skills for Claude Code and compatible agents. Each skill is a self-contained folder in skills/ with a SKILL.md at its root.
Project constraints
- Bundled scripts use Bash and Python 3 (stdlib only); the org-wide Bun preference does not apply. Skills declare external CLI prerequisites in
SKILL.mdand fail with exact install and rerun guidance when an applicable tool is missing. They do not resolve tool packages at runtime. Skills that create projects or packages may install the requested dependencies. - Skills install independently. Do not assume sibling skills are installed; follow
.agents/rules/repo-conventions.mdfor cross-skill cooperation. User documentation lives in the root README, with no per-skill root README.
Rule index
- Authoring or evaluating a skill: read
.agents/rules/skill-authoring.mdand use the official Anthropicskill-creator. - Frontmatter, layout, or portability: read
.agents/rules/agentskills-spec.md; read.agents/rules/claude-code-skills.mdfor Claude Code extensions and substitutions. - Flags, output paths, installation, marketplace, or CI changes: read
.agents/rules/repo-conventions.md. - Adding, changing, or removing tests: read Testing for regression value, coverage preservation, and execution limits.
- Shared skill instructions: edit the owning
.agents/rules/skill-{prose,label-hygiene,execution-discipline,adversarial-verification}-rules.mdblock or.agents/rules/skill-quality-lens-rules.md. These sources declare their recipients; the sync script updates the bundled copies.
Validation
- Run
python3 -m unittest discover tests/before reporting done and before commit. Report failures, skips and their reasons, and unavailable external checks; passing tests alone do not establish skill behavior or host compatibility. - After editing a canonical
skill-*-rules.mdblock, runpython3 scripts/sync_writing_rules.pyand review the propagated diff.
Release
Use branch main. Release notes belong only in the gh release create body; there is no CHANGELOG.md. Versions live only in git tags and .claude-plugin/marketplace.json metadata.version. There is no root package.json or pnpm version step.
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.