agentleFS
Sign inStart free
jaredstarnerjaredstarner/agent-skills/AGENTS.md

agent-skills

AGENTS.md0 starsChanged 27 days agoMITView on GitHub

Shareable skills in the Agent Skills format. Each top-level folder is one skill: a SKILL.md plus optional references/ and scripts/. CLAUDE.md is a symlink to this file; edit AGENTS.md. research/ (source material used to write skills) and dist/ (packaged zips) stay local and are gitignored. Never commit transcripts or extracts.

  • 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 repository6 other files it gives its agents.

The file

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

6 rules. About 0 are common, 3 are distinctive, and 3 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.

  • 3

    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
  1. Distinctive:

    Never push to main; work on a branch and open a pull request into main

    line 7
  2. Distinctive:

    Keep commits small and focused.

    line 9
  3. Very specific:

    Write commit messages as Conventional Commits, scoped to the skill where it fits (feat(website): ...).

    line 10
  4. Very specific:

    Run python .github/scripts/validate_skills.py before pushing.

    line 12
  5. Very specific:

    Every PR needs the owner's approval (.github/CODEOWNERS).

    line 13
  6. Distinctive:

    No em-dashes in any file

    line 25

≈ “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-skills

Shareable skills in the Agent Skills format. Each top-level folder is one skill: a SKILL.md plus optional references/ and scripts/. CLAUDE.md is a symlink to this file; edit AGENTS.md.

Workflow

  • Never push to main; work on a branch and open a pull request into main
  • Sign every commit; the main ruleset rejects unsigned commits
  • Keep commits small and focused; PRs merge with a merge commit, so each commit lands on main
  • Write commit messages as Conventional Commits, scoped to the skill where it fits (feat(website): ...). There are no releases; main is what users install
  • Name branches <type>/<short-description>, using a Conventional Commit type
  • Run python .github/scripts/validate_skills.py before pushing; CI runs it as the required validate-skills check
  • Every PR needs the owner's approval (.github/CODEOWNERS). The owner merges their own PRs with the admin bypass once checks pass
  • Agents stop for a local review before pushing a branch or opening a PR

Skill rules

  • A skill holds its domain method only. Personal working preferences (tone, approval gates, formatting habits) belong in the user's own agent instructions, not in a skill
  • Tool-agnostic by default: name a common default (for example Jira for tickets, gh for the code host) and work with another tool when the user names one. Map the concepts the skill needs to that tool and confirm before proceeding; if no tool exists for it, stop
  • Nothing personal or employer-specific: no names, emails, local paths, internal hosts, or real ticket keys. Use generic examples
  • Frontmatter keeps to name, description, and license: MIT. name matches the folder. description stays under 1024 characters and says what the skill does and when to load it; write it as a >- block or quote it, since a bare colon breaks the YAML
  • Refer to agent actions generically ("ask the user"), not by one agent's tool names
  • Requirements (runtimes, CLIs, accounts) go in a "Requirements" section of the body
  • Scripts state their runtime in --help and the README; avoid dependencies where possible
  • No em-dashes in any file
  • Adding a skill: add its README table row and add it to the skill dropdown in .github/ISSUE_TEMPLATE/skill-bug.yml and skill-change.yml; the validator checks all three

Not in the repo

research/ (source material used to write skills) and dist/ (packaged zips) stay local and are gitignored. Never commit transcripts or extracts.

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