claude-skills
AGENTS.md0 starsChanged 5 months agoMITView on GitHub
Guidance for AI coding agents (Claude Code, Cursor, and others) working with this repo. Humans: start with README.md. This file follows the agents.md convention. A curated collection of self-contained Claude Code skills. Each skill lives in skills/<name>/ and is independent — no skill depends on another, and none depends on anything outside its own folder. You can copy any one folder into a ~/.claude/skills/ directory and it works on its own. - catalog.json — the machine-readable index. Parse it to…
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 31,700 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.
- llms.txtllms.txt
- annotateSkill
- debug-loop-breakerSkill
- deploy-checklistSkill
- goalSkill
- growth-playbookSkill
- security-scannerSkill
- seo-auditSkill
- svg-diagramSkill
- tdd-disciplineSkill
- teaching-modeSkill
- verify-frontend-changeSkill
- SKILL_NAMESkill
The file
We split this AGENTS.md into the rules it gives an agent
5 rules. About 0 are common, 1 is distinctive, and 4 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.
4
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 24
Read its skills/<name>/SKILL.md and follow the procedure there.
- Very specific: line 25
If it references files (references/…, scripts/…), they are inside that same skill folder.
- Very specific: line 28
If the skill ships scripts, run them with the command shown in its SKILL.md.
- Very specific: line 41
No personalization: No author names, private domains, private project names, or machine-local paths.
- Distinctive: line 60
Never edit the catalog table in README.md by hand — it is generated.
≈ “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
AGENTS.md
Guidance for AI coding agents (Claude Code, Cursor, and others) working with this
repo. Humans: start with README.md. This file follows the
agents.md convention.
What this repo is
A curated collection of self-contained Claude Code skills. Each skill lives in
skills/<name>/ and is independent — no skill depends on another, and none depends
on anything outside its own folder. You can copy any one folder into a
~/.claude/skills/ directory and it works on its own.
Discover what's here
catalog.json— the machine-readable index. Parse it to list every skill with its name, description, tags, maturity, and path. It is generated from the skills themselves, so it never drifts.skills/<name>/SKILL.md— the skill's own instructions. The frontmatterdescriptionis the routing signal (when to trigger); the body is the procedure.
Use a skill
- Read its
skills/<name>/SKILL.mdand follow the procedure there. - If it references files (
references/…,scripts/…), they are inside that same skill folder. Paths in a SKILL.md are relative to the skill folder — never an absolute machine path. - If the skill ships scripts, run them with the command shown in its SKILL.md.
Install a skill (for the user)
cp -r skills/<name> ~/.claude/skills/That's the whole install. No build step. Skills are plain Markdown (+ optional scripts/assets).
The skill contract — invariants every skill in this repo holds
- Self-contained. Everything the skill needs is inside its folder.
- No personalization. No author names, private domains, private project names,
or machine-local paths. Enforced by
tools/check-skill.sh. - Portable paths. References are relative to the skill folder.
- A
SKILL.mdwith frontmatter (name,description) is mandatory; aREADME.md(human) andmeta.json(tags/maturity) are expected. - Gotchas live in the skill. Hard-won edge cases go in the skill's own
## Gotchas, not in tribal memory.
Editing or adding a skill
See CONTRIBUTING.md. The short version:
tools/new-skill.sh <name> # scaffold from _template/
# ...write SKILL.md / README.md / meta.json...
tools/check-skill.sh skills/<name> # MUST pass — no secrets, no personalization
python3 tools/build-catalog.py # regenerate catalog.json + the README tableNever edit the catalog table in README.md by hand — it is generated.
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.