agentleFS
Sign inStart free
santanusetusantanusetu/claude-skills/CLAUDE.md

claude-skills

CLAUDE.md0 starsChanged 2 months agoMITView on GitHub

A public collection of Claude Code skills, installable as plugins. Everything here is public, so read the privacy rule first. One skill per folder. A skill's folder holds only what gets installed, since tests and fixtures in skills/ would be copied onto every user's machine. Repo tooling lives under .github/ so the top level stays clean. Then fill both in. 2. Write the description so it says what the skill does and when to use it. That text is what…

The file

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

6 rules. About 0 are common, 2 are distinctive, and 4 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.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Examples, fixtures, evals and screenshots are fictional: Never paste real transcripts, names, employers, emails, file paths or session IDs, …

    line 21
  2. Very specific:

    Write the description so it says what the skill does and when to use it.

    line 34
  3. Very specific:

    Put deterministic work in scripts/, with Python stdlib only if possible, and let Claude do the judgement.

    line 35
  4. Very specific:

    Add fictional fixtures and unittest tests in tests/<name>/.

    line 36
  5. Very specific:

    Add a plugin entry to .claude-plugin/marketplace.json, and the skill to README.md: a line under Skills plus its own section below.

    line 37
  6. Distinctive:

    Commit as the repo owner.

    line 44

≈ “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 — conventions

A public collection of Claude Code skills, installable as plugins. Everything here is public, so read the privacy rule first.

Layout

skills/<name>/                    the skill: SKILL.md + scripts/ + assets/ (what users install)
tests/<name>/                     unit tests, fictional fixtures, evals.json, screenshot builder
template/                         copy this to start a skill (SKILL.md, scripts/, tests/)
.claude-plugin/marketplace.json   one plugin entry per skill
.github/workflows/                CI: every skill's tests + the privacy check
.github/scripts/privacy_check.py  blocks commits that contain private terms
.github/media/                    images used by README.md

One skill per folder. A skill's folder holds only what gets installed, since tests and fixtures in skills/ would be copied onto every user's machine. Repo tooling lives under .github/ so the top level stays clean.

Privacy rule

  • Examples, fixtures, evals and screenshots are fictional. Never paste real transcripts, names, employers, emails, file paths or session IDs, even to "just test something".
  • Before every commit, run python3 .github/scripts/privacy_check.py. It reads your own terms from .private-patterns at the repo root, which is gitignored and never committed, plus built-in checks for home-folder paths and emails. Start from .github/private-patterns.example.
  • *-workspace/ folders from skill-creator are gitignored, because they contain real run outputs. Keep them outside the repo anyway.

Adding a skill

  1. Copy the template into place:
    mkdir -p skills/<name> tests/<name>
cp -R template/SKILL.md template/scripts skills/<name>/
cp template/tests/* tests/<name>/
    Then fill both in.
  2. Write the description so it says what the skill does and when to use it. That text is what makes Claude pick the skill.
  3. Put deterministic work in scripts/, with Python stdlib only if possible, and let Claude do the judgement.
  4. Add fictional fixtures and unittest tests in tests/<name>/.
  5. Add a plugin entry to .claude-plugin/marketplace.json, and the skill to README.md: a line under Skills plus its own section below.
  6. Run the checks:
    python3 -m unittest discover -s tests/<name>
python3 .github/scripts/privacy_check.py
claude plugin validate .
  7. Commit as the repo owner.

Style

  • Keep SKILL.md under ~200 lines. Explain why rather than stacking MUSTs.
  • The README screenshots for a skill must come from its fixtures, so readers can reproduce them. Store them in .github/media/<name>-*.png.

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