agentleFS
Sign inStart free
otomata-techotomata-tech/claude-skills/CLAUDE.md

claude-skills

CLAUDE.md1 starsChanged 5 months agoMITView on GitHub

Repository of generic, reusable skills for Claude Code. Each skill is a directory with a SKILL.md; users install via symlink into ~/.claude/skills/. See README.md for the user-facing skill list and other skill sources. One skill = skills/<slug>/ with a required SKILL.md (+ optional scripts/). scripts/build-api.sh aggregates every SKILL.md frontmatter into dist/api/ (a flat skills.json + one skills/<slug>.json per skill). name is what users type as /<name> — it can differ from the directory name (e.g. skills/session/ exposes itself as session-lookup).

The file

We split this CLAUDE.md into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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
  1. Distinctive:

    No tests, no lint — skills are markdown.

    line 33
  2. Distinctive:

    No legacy fallbacks in skill bodies — if a skill is renamed, update references rather than aliasing.

    line 41
  3. Very specific:

    Keep SKILL.md under ~200 lines unless the skill genuinely needs more — long instructions mean the model picks up less of it.

    line 43
  4. Very specific:

    Add : skills/<slug>/SKILL.md + row in README's ## Skills table + symlink into ~/.claude/skills/<slug> to test locally + …

    line 48

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

claude-skills

Repository of generic, reusable skills for Claude Code. Each skill is a directory with a SKILL.md; users install via symlink into ~/.claude/skills/. See README.md for the user-facing skill list and other skill sources.

Structure

One skill = skills/<slug>/ with a required SKILL.md (+ optional scripts/). scripts/build-api.sh aggregates every SKILL.md frontmatter into dist/api/ (a flat skills.json + one skills/<slug>.json per skill).

SKILL.md frontmatter

---
name: <slug>                   # Falls back to dir name if absent
description: <one line>        # Used by the model to decide relevance
disable-model-invocation: false
argument-hint: "[arg1|arg2]"   # Optional, displayed in autocomplete
---

name is what users type as /<name> — it can differ from the directory name (e.g. skills/session/ exposes itself as session-lookup).

Commands

./scripts/build-api.sh   # Regenerate dist/api/ JSON from all SKILL.md

No tests, no lint — skills are markdown.

Conventions

  • Slug = directory name when possible; a different name: only when there's a meaningful reason (fs slug generic but the skill needs a specific public name).
  • Description must be specific enough for the model to decide whether to invoke — avoid vague descriptions like "helper for stuff".
  • No legacy fallbacks in skill bodies — if a skill is renamed, update references rather than aliasing.
  • Keep SKILL.md under ~200 lines unless the skill genuinely needs more — long instructions mean the model picks up less of it.

Adding / renaming a skill

  • Add : skills/<slug>/SKILL.md + row in README's ## Skills table + symlink into ~/.claude/skills/<slug> to test locally + ./scripts/build-api.sh if the JSON API is consumed externally.
  • Rename : git mv skills/<old> skills/<new>, update name: in SKILL.md + README, replace the symlink.

Other repos contribute skills the same way (/data/oto/skills/, /data/howto/skills/, etc.) — see README's "Other skill sources" table.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free