agentleFS
Sign inStart free
birdsighbirdsigh/Skills/AGENTS.md

Skills

AGENTS.md3 starsChanged 3 months agoMITView on GitHub

This repository contains small, self-contained Codex skills, one directory per skill. When creating or editing a skill in collaboration with the user, always work on the version in this repository. Never edit the copy installed in the currently active agent. Installed skill copies are runtime artifacts and may be stale or replaced. Treat <repo>/<skill-name>/ as the editable source. Installation or cache refresh is a separate, explicit step. There is no repo-wide build, test, lint, or typecheck command. Review the changed…

  • 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 repository9 other files it gives its agents.

The file

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

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

  • 5

    Very specific

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

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

    Read the target skill's complete SKILL.md before changing it.

    line 13
  2. Very specific:

    Keep each skill self-contained.

    line 14
  3. Distinctive:

    Follow the existing direct, procedural Markdown style.

    line 16
  4. Very specific:

    Do not modify unrelated skills or user-owned worktree changes.

    line 17
  5. Very specific:

    When committing changes in this repository, do not use the commit-and-push skill.

    line 25
  6. Very specific:

    Keep skills focused and self-contained.

    line 43
  7. Distinctive:

    Add bad/good example pairs here as recurring agent mistakes are observed.

    line 48

≈ “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

NiceSkills agent instructions

This repository contains small, self-contained Codex skills, one directory per skill.

Source of truth

When creating or editing a skill in collaboration with the user, always work on the version in this repository. Never edit the copy installed in the currently active agent.

Installed skill copies are runtime artifacts and may be stale or replaced. Treat <repo>/<skill-name>/ as the editable source. Installation or cache refresh is a separate, explicit step.

Working in this repository

  • Read the target skill's complete SKILL.md before changing it.
  • Keep each skill self-contained. Put supporting material under that skill's scripts/, references/, or assets/ directory.
  • Preserve YAML frontmatter with a concise name and trigger-focused description.
  • Follow the existing direct, procedural Markdown style. Keep instructions specific enough for an agent to execute.
  • Do not modify unrelated skills or user-owned worktree changes.

Verification

There is no repo-wide build, test, lint, or typecheck command. Review the changed Markdown and diff directly. For a skill with scripts, run the smallest relevant script-level check available.

Git

When committing changes in this repository, do not use the commit-and-push skill. This prohibition applies to both the repository copy and any copy installed in the current agent harness. Use direct Git commands and follow the user's explicit commit and push instructions instead.

Repository map

  • <skill-name>/SKILL.md: canonical skill definition and workflow.
  • <skill-name>/scripts/: executable helpers owned by that skill.
  • <skill-name>/references/: supporting detail loaded only when relevant.
  • README.md: human-facing catalogue of the skills.

Glossary

  • Repository copy: canonical editable version of a skill in this project.
  • Installed copy: version loaded into an active agent environment; never the editing target.
  • Skill: a directory whose SKILL.md defines triggers and operating instructions.

Never compromise

  • Edit the repository copy, never the active agent's installed copy.
  • Keep skills focused and self-contained.
  • Preserve unrelated worktree changes.

Failure modes

Add bad/good example pairs here as recurring agent mistakes are observed.

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