agentleFS
Sign inStart free
grezxunegrezxune/ai-tools/AGENTS.md

ai-tools

AGENTS.md0 starsChanged 18 days agoMITView on GitHub

The rules and skills our coding agents load, shared between Claude Code and Codex. scripts/install.sh links them into the home directory, so an edit here changes every later session on that machine.

  • 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 repository13 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, 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. Very specific:

    Before writing or editing a skill or a rules file, read skills/writing-for-agents/SKILL.md.

    line 12
  2. Very specific:

    After changing a script, run bash -n on it, then run it once.

    line 17
  3. Distinctive:

    When a skill is added, removed, or changes how it loads, update the tables in README.md.

    line 18
  4. Very specific:

    After scripts/update-skills.sh, read git diff skills before committing.

    line 23
  5. Distinctive:

    No AI-assistant attribution in commit messages, PR descriptions, or code comments.

    line 28
  6. Very specific:

    Update CHANGELOG.md for meaningful changes.

    line 30

≈ “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

ai-tools

The rules and skills our coding agents load, shared between Claude Code and Codex. scripts/install.sh links them into the home directory, so an edit here changes every later session on that machine.

Project notes

  • rules/AGENTS.md loads in every session. Each line costs context on every turn, so a rule goes there only when it applies in every repo.
  • A skill's description also loads in every session. Write what the skill is, then the cases that should trigger it, and nothing else.
  • A skill that only runs by hand sets disable-model-invocation: true in its frontmatter, and policy.allow_implicit_invocation: false in agents/openai.yaml for Codex.
  • Skills listed in skills/SOURCES are vendored and match upstream byte for byte. Change them with scripts/update-skills.sh, never by hand. Local rules go in another skill.
  • Skills point at each other by relative path, such as ../prd/SKILL.md, because the folder is linked into more than one place.
  • Before writing or editing a skill or a rules file, read skills/writing-for-agents/SKILL.md.

Code

  • Scripts are bash, start with set -euo pipefail, and are safe to re-run.
  • After changing a script, run bash -n on it, then run it once.
  • When a skill is added, removed, or changes how it loads, update the tables in README.md.

Security

  • This repo is public. Keep secrets, tokens, client names, and machine-specific paths out of every file.
  • After scripts/update-skills.sh, read git diff skills before committing. A skill is instructions an agent will follow.

Git

  • Conventional Commits: feat:, fix:, chore:, refactor:.
  • No AI-assistant attribution in commit messages, PR descriptions, or code comments.
  • Branch as feature/{name}. Squash before merging.
  • Update CHANGELOG.md for meaningful changes.

Docs

  • Writing: plain prose in docs, PR descriptions, and skill text. No em dashes, no puffery, no filler. Say what a thing does.

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