agentleFS
Sign inStart free
dawn840705dawn840705/codex-code-studios/AGENTS.md

codex-code-studios

AGENTS.md17 starsChanged 2 months agoMITView on GitHub

This repository is the source package for the codex-code-studios Codex plugin. This studio ships twice: codex-code-studios (Codex, $skill-name) and the sibling claude-code-studios (Claude Code, /skill-name). A user project may be worked by one alone or by both at once, and that decides whether this plugin acts at all. production/runtime.txt in the user's project is the source of truth — one line: codex (full studio, default) · claude (stand down; the sibling owns the work) · split (work only the partition…

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
  1. Very specific:

    Use $skill-name when referring to a Codex skill invocation.

    line 34
  2. Very specific:

    Keep hook stdout valid for the matching Codex hook event.

    line 36
  3. Distinctive:

    Run the plugin validator, every skill validator, the skill linter, shell syntax checks, and the full test suite before release.

    line 37
  4. Distinctive:

    When changing asset tracking, ignore rules, or cross-machine handoff workflows, read Artifact Organization Rules, including the shared …

    line 38

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 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

Code Studios contributor guide

This repository is the source package for the codex-code-studios Codex plugin.

Source of truth

  • .codex-plugin/plugin.json defines the Codex plugin.
  • .agents/plugins/marketplace.json defines the local/shareable marketplace entry.
  • skills/*/SKILL.md contains user-facing workflows. Keep every skill valid under the Codex skill format.
  • skills/studio-orchestrator/references/roles/*.md contains specialist personas. They are references loaded by the orchestrator, not independently installed Codex agents.
  • hooks/hooks.json and hooks/*.sh implement Codex lifecycle hooks.
  • rules/*.md, docs/, and templates/ are plugin resources. They are not Codex command-approval rule files.

Runtime modes

This studio ships twice: codex-code-studios (Codex, $skill-name) and the sibling claude-code-studios (Claude Code, /skill-name). A user project may be worked by one alone or by both at once, and that decides whether this plugin acts at all.

production/runtime.txt in the user's project is the source of truth — one line: codex (full studio, default) · claude (stand down; the sibling owns the work) · split (work only the partition this runtime owns). Absent means codex. Never infer the mode from a CLAUDE.md or .claude/ sighting.

In split mode, write ownership is partitioned by path before work starts. The two runtimes share no session, context or tool log, so an overwrite is invisible to both and the losing side still reports success. Full rule: rules/runtime-modes.md.

Change rules

  1. Preserve existing public skill names unless a migration explicitly documents a rename.
  2. Use $skill-name when referring to a Codex skill invocation.
  3. Resolve bundled resources relative to the invoking SKILL.md or PLUGIN_ROOT; reserve .codex/ for state stored in the user's project.
  4. Keep hook stdout valid for the matching Codex hook event. Advisory PostToolUse feedback belongs in hookSpecificOutput.additionalContext.
  5. Run the plugin validator, every skill validator, the skill linter, shell syntax checks, and the full test suite before release.
  6. When changing asset tracking, ignore rules, or cross-machine handoff workflows, read Artifact Organization Rules, including the shared asset recovery requirements.

The legacy .claude-plugin/ manifest and CLAUDE.md remain only as a transition compatibility layer for the 0.7 release. Do not make them the source of truth for new Codex behavior.

프로젝트에서 확인한 방침의 지속 반영

  • 공유 기억·결정·인계는 work-records에 따라 저장소 파일에 둔다. 인계 지시문의 파일 링크를 먼저 읽고 개인 에이전트 메모리를 정본으로 인용하지 않는다.
  • 프로젝트 진행 중 확정된 공통 운영 원칙과 검증된 재발 방지 절차는 이 원본 저장소에 계속 반영한다. 사용자 직접 지시 근거: 2026-09-07, 새 코드 스튜디오 프로젝트에 앞으로 가져갈 방침을 업데이트하라는 요청.
  • 반영 시 공통 방침 갱신 절차를 읽고 따른다. 기존 규칙·스킬·템플릿의 실제 적용 위치를 갱신하고 중복 정본을 만들지 않는다.
  • 특정 제품의 기획·수치·개인 경로·역할 계약은 공통 기본값으로 승격하지 않는다. 원본 변경과 설치된 플러그인 적용 상태를 따로 보고한다.

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