agentleFS
Sign inStart free
XakkiXakki/ai-agents-skills/AGENTS.md

ai-agents-skills

AGENTS.md6 starsChanged 3 months agoMITView on GitHub

This repository ships five integrations from the same source tree: Claude Code, Codex, Cursor Agent CLI/IDE, Hermes Agent, and Prime Agent: - .claude-plugin/plugin.json + .claude-plugin/marketplace.json — Claude Code manifest and marketplace entry. - .codex-plugin/plugin.json — Codex plugin manifest. - .cursor-plugin/plugin.json + .cursor-plugin/marketplace.json — Cursor Agent CLI/IDE manifest and marketplace entry. - plugin.yaml + __init__.py — Hermes manifest and Python registration adapter. It registers the shared skills read-only under the ai-agents-skills: namespace and maps compatible lifecycle hooks. - package.json + extensions/prime-adapter.ts…

The file

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

3 rules. About 0 are common, 2 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When changing shared behavior (a skill, an agent, a hook script), preserve

    line 37
  2. Very specific:

    Run pytest -q tests/test_hermes_plugin.py and an isolated hermes plugins install file://...

    line 47
  3. Distinctive:

    All authored repository text (skills, docs, manifests, commit messages) is English.

    line 49

≈ “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

AGENTS.md

This repository ships five integrations from the same source tree: Claude Code, Codex, Cursor Agent CLI/IDE, Hermes Agent, and Prime Agent:

  • .claude-plugin/plugin.json + .claude-plugin/marketplace.json — Claude Code manifest and marketplace entry.
  • .codex-plugin/plugin.json — Codex plugin manifest.
  • .cursor-plugin/plugin.json + .cursor-plugin/marketplace.json — Cursor Agent CLI/IDE manifest and marketplace entry.
  • plugin.yaml + __init__.py — Hermes manifest and Python registration adapter. It registers the shared skills read-only under the ai-agents-skills: namespace and maps compatible lifecycle hooks.
  • package.json + extensions/prime-adapter.ts — Prime Agent capability-package manifest and lifecycle adapter. It exposes skills/ and agents/ through the Prime package resource API, then maps only lifecycle events Prime supports.

Conventions

  • Shared skills live under skills/ and are used by all four integrations — don't fork or duplicate a skill per runtime.
  • Runtime-specific lifecycle hook maps stay separate: hooks/hooks.json (Claude Code), hooks/codex-hooks.json (Codex, declared explicitly in .codex-plugin/plugin.json), and hooks/cursor-hooks.json (Cursor — routes every event through hooks/cursor-adapter.sh, which normalizes Cursor's payload field names into the shape the shared tg-*.sh/abbr-inject.sh scripts already expect). Do not merge them, and put any Cursor-specific translation logic in the adapter, not in the shared scripts. Cursor has no Notification/PermissionRequest equivalent — the "needs attention" ping is intentionally not wired for Cursor; see docs/superpowers/specs/2026-07-31-cursor-plugin-design.md.
  • Hermes lifecycle translation belongs in the root __init__.py; shared shell hooks continue to consume the normalized Claude-shaped JSON payload. Hermes has no permission/idle notification event, so that ping degrades safely just as it does under Cursor.
  • When changing shared behavior (a skill, an agent, a hook script), preserve all five integrations — verify the change still works for Claude Code, Codex, Cursor, Hermes, and Prime Agent, and update every manifest the change affects metadata for (e.g. version).
  • Codex hooks must be declared explicitly as ./hooks/codex-hooks.json so Codex does not auto-discover the Claude Code-specific hooks/hooks.json. Plugin hooks require the user's Codex trust review before they can execute.
  • Script-bearing skills must resolve the root through AI_AGENTS_SKILLS_ROOT, with Claude/Codex/Cursor root fallbacks. Hermes sets AI_AGENTS_SKILLS_ROOT during plugin registration.
  • Run pytest -q tests/test_hermes_plugin.py and an isolated hermes plugins install file://... --enable smoke test for Hermes changes.
  • All authored repository text (skills, docs, manifests, commit messages) is English.
  • See README.md for install/enable/verify steps for each runtime.

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