rlajousrlajous/claude-code-commands/AGENTS.md

claude-code-commands

AGENTS.md30 starsChanged 3 months agoMITView on GitHub

This repository ships the same workflow skills to Claude Code and Codex. Keep shared instructions host-neutral and preserve compatibility for both runtimes. - skills/*/SKILL.md contains the 21 shared workflow skills and their canonical bundled resources. - agents/*.md contains the eight canonical agent instructions. - .codex/agents/*.toml is generated by node scripts/generate-codex-agents.mjs; never edit those files by hand. - .git-workflow/config.yaml and .git-workflow/ state files are canonical. Read matching legacy .claude/ files only as backward-compatible fallbacks. - hooks/review-commit.sh is shared. The notifications skill…

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 repository23 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- change-briefSkill
- clean-goneSkill
- close-customer-loopSkill
- commitSkill
- finishSkill
- notificationsSkill
- plan-qaSkill
- plan-tddSkill
- release-notesSkill
- releaseSkill
- review-requestSkill
- reviewSkill
- review-watchSkill
- rfcSkill
- setupSkill
- standupSkill
- start-qaSkill
- startSkill
- statusSkill
- syncSkill
- tddSkill
- updateSkill

The file

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

4 rules. About 0 are common, 2 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Document Claude invocation as /skill-name and Codex invocation as $skill-name.
line 17
Very specific:
Keep deterministic helpers in the owning skill's scripts/ or assets/ directory.
line 23
Distinctive:
Before submitting changes, run:
line 31
Distinctive:
If canonical agent instructions changed, regenerate and commit their Codex definitions:
line 37

≈ “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

Git Workflow contributor guidance

This repository ships the same workflow skills to Claude Code and Codex. Keep shared instructions host-neutral and preserve compatibility for both runtimes.

Source of truth

skills/*/SKILL.md contains the 21 shared workflow skills and their canonical bundled resources.
agents/*.md contains the eight canonical agent instructions.
.codex/agents/*.toml is generated by node scripts/generate-codex-agents.mjs; never edit those files by hand.
.git-workflow/config.yaml and .git-workflow/ state files are canonical. Read matching legacy .claude/ files only as backward-compatible fallbacks.
hooks/review-commit.sh is shared. The notifications skill owns the generic notifier, unified GitHub daemon, configuration parser, and Stop-hook adapter. Host-specific registration lives in hooks/claude-hooks.json and hooks/hooks.json.

Runtime conventions

Document Claude invocation as /skill-name and Codex invocation as $skill-name.
Describe actions such as asking the user, searching the web, or delegating without naming host-specific tools.
Resolve bundled resources from the absolute path of the loaded SKILL.md. Derive the package root from that skill directory only for package-wide assets. Accept PLUGIN_ROOT, then CLAUDE_PLUGIN_ROOT, only as verified compatibility fallbacks when a host cannot expose the loaded path.
Keep deterministic helpers in the owning skill's scripts/ or assets/ directory. Root scripts/* entries may be compatibility wrappers, but skills must never silently fall back to /scripts or use a root wrapper as their canonical implementation.
Preserve Claude-specific frontmatter where Claude requires it. Codex uses name and description from each skill plus the generated TOML agent definitions.
Routine edits stay in the main session. Only review, release validation/version comparison, and sufficiently large QA execution workflows delegate to named agents.

Validation

Before submitting changes, run:

bash scripts/validate.sh

If canonical agent instructions changed, regenerate and commit their Codex definitions:

node scripts/generate-codex-agents.mjs

See docs/SUBAGENTS.md for the user-facing agent catalog and CLAUDE.md for Claude-specific contributor notes.

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