agentleFS
Sign inStart free
lggarrisonlggarrison/agent-skills/.claude/skills/adding-a-skill/SKILL.md

adding-a-skill

Skill0 starsChanged 23 days agoMITView on GitHub

How to add a new skill to this repository, or change an existing one, so it fits INTENTION.md and passes the checks. Use when asked to create, rename, split, or substantially change a skill in skills/. This skill is for this repository only and isn't published.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. BuilderIO/skills★ 4.7ksame name
    2. stellar/stellar-core★ 3.3ksame name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository8 other files it gives its agents.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 4 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If a skill-creator skill is available, use it for drafting and testing the skill's content.

    line 14
  2. Very specific:

    Read INTENTION.md first. A skill belongs in skills/ only if it:

    line 19
  3. Very specific:

    If it only makes sense in one project, it belongs in that project's CLAUDE.md, AGENTS.md, or local skills.

    line 26
  4. Distinctive:

    If it needs a script or CLI, the tooling goes in its own package and the skill tells agents how to use it.

    line 29
  5. Very specific:

    Create skills/<name>/SKILL.md.

    line 34
  6. Distinctive:

    Keep it short and specific: rules an agent can follow and a reviewer can check.

    line 40
  7. Distinctive:

    No project names, paths, package scripts, internal tool names, secrets, or personal data.

    line 42
  8. Distinctive:

    Add a row to the skills table in README.md: the skill's link and when to use it.

    line 49
  9. Very specific:

    Add a line under Unreleased in CHANGELOG.md, under Added or Changed.

    line 52
  10. Very specific:

    Update its description if what it does or when to use it changed, and its README row if its use changed.

    line 57
  11. Very specific:

    When you rename a heading, search the other skills and docs for links to its old anchor (grep -rn '#old-anchor' .).

    line 59
  12. Very specific:

    When renaming a skill, rename its folder, its name, its README row, and its symlink, and update every link to it.

    line 62
  13. Very specific:

    Run npm run check and fix everything it reports.

    line 67

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
SKILL.md

Adding or changing a skill in this repository

This repository publishes general-purpose skills from skills/. This skill lives in .claude/skills/ as a real folder rather than a link, so it's used here and never installed into other projects. INTENTION.md and CONTRIBUTING.md are the source of truth. When they disagree with this skill, they win.

If a skill-creator skill is available, use it for drafting and testing the skill's content. This skill covers what this repository needs on top of that.

1. Check that it belongs

Read INTENTION.md first. A skill belongs in skills/ only if it:

  • works in any repository and describes a way of working, not a particular project;
  • is plain Markdown any agent can follow, with no dependency on one vendor's features unless it says so and offers a fallback;
  • is the author's own work, with nothing taken from an employer or client.

If it only makes sense in one project, it belongs in that project's CLAUDE.md, AGENTS.md, or local skills. If it only applies to this repository, put it in .claude/skills/<name>/ as a real folder, like this one, and skip steps 3 and 4 below. If it needs a script or CLI, the tooling goes in its own package and the skill tells agents how to use it.

2. Write it

  • Create skills/<name>/SKILL.md. <name> is lowercase letters, digits, and hyphens, at most 64 characters. The frontmatter name must match the folder.
  • The description says what the skill does and when to use it, in at most 1,024 characters. Agents read only this when deciding whether to load the skill, so name the requests that should trigger it.
  • Open by deferring to the repository's own rules. Look at how the existing skills do it.
  • Keep it short and specific: rules an agent can follow and a reviewer can check. Cut anything an agent would do correctly without being told.
  • No project names, paths, package scripts, internal tool names, secrets, or personal data.
  • Link to sibling skills with relative paths (../pr-review/SKILL.md). If two skills link to each other, say in the README that they're installed together.

3. Wire it up

  • Add a row to the skills table in README.md: the skill's link and when to use it.
  • Link it for Claude Code: ln -s ../../skills/<name> .claude/skills/<name>.
  • Add a line under Unreleased in CHANGELOG.md, under Added or Changed.

4. When changing an existing skill

  • Update its description if what it does or when to use it changed, and its README row if its use changed.
  • When you rename a heading, search the other skills and docs for links to its old anchor (grep -rn '#old-anchor' .). npm run check checks that linked files exist, but not anchors.
  • When renaming a skill, rename its folder, its name, its README row, and its symlink, and update every link to it.

5. Check and open the PR

  • Run npm run check and fix everything it reports. Use npm run format for formatting.
  • Try the skill in a real repository with at least one agent, and say in the PR which agent you used.
  • Work through the checklist in INTENTION.md, then open the PR with the github skill. The repository's PR template has the same checklist.

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