agentleFS
Sign inStart free
patrykmurzynpatrykmurzyn/skills/AGENTS.md

skills

AGENTS.md0 starsChanged 17 days agoMITView on GitHub

This repo holds public Agent Skills. README.md covers the layout and install.

  • 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 31,542 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 repository9 other files it gives its agents.

The file

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

5 rules. About 0 are common, 3 are distinctive, and 2 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.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When a skill in a group is added, renamed or removed, update the group's router skill in the same change.

    line 7
  2. Distinctive:

    Every change to a skill names the failure it fixes (a baseline run, a failed eval or a user correction) in the commit message.

    line 8
  3. Very specific:

    Write skills and docs in English.

    line 12
  4. Very specific:

    After any change, scripts/validate.sh must pass.

    line 13
  5. Distinctive:

    Don't push, tag or publish without the owner's explicit yes.

    line 18

≈ “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.

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

Agent instructions

This repo holds public Agent Skills. README.md covers the layout and install.

  • Create or change a skill with the skill-authoring skill: find friction (optional), then scope, write and test. New skills never skip scoping. Friction findings from friction-mining or friction-probe are input to scoping, never a skill by themselves. Vet third-party skills with skill-vetting, and review installed ones with skill-review.
  • One job per skill. Related skills share a group folder, skills/<group>/. Names are unique across the repo.
  • When a skill in a group is added, renamed or removed, update the group's router skill in the same change.
  • Every change to a skill names the failure it fixes (a baseline run, a failed eval or a user correction) in the commit message.
  • Frontmatter: name, description, license: MIT, and metadata with author: patrykmurzyn, version (quoted semver) and status (experimental, stable or deprecated). Nothing else: claude.ai rejects other fields, and scripts/validate.sh does too.
  • Bump metadata.version on every change to a skill: patch for wording, minor for new behaviour, major for a changed contract. Set stable only after the skill's evals have run and passed. A deprecated skill is removed in the next release.
  • Skills refer to other skills by name in backticks, never by path.
  • Write skills and docs in English.
  • After any change, scripts/validate.sh must pass. Then run uv run scripts/catalog.py and add a line to CHANGELOG.md under Unreleased.
  • Fixtures that are meant to fail live in evals/files/; the tooling skips that folder.
  • Eval workspaces (*-workspace/) stay out of git.
  • Install third-party skills with their own tool after vetting; don't copy them here. A modified fork keeps its license and names its origin in metadata.source.
  • This repo is public: no secrets, personal data or machine-specific paths.
  • Don't push, tag or publish without the owner's explicit yes.

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