agentleFS
bazaarjapanbazaarjapan/codex-autoloop-template/.agents/skills/autoloop-setup/SKILL.md

autoloop-setup

Skill0 starsChanged 3 months agoMITView on GitHub

Initialize a copied Codex AutoLoop repository for ChatGPT subscription use. Use when the user says "初期設定して", "セットアップして", "initialize this repository", or asks to make AutoLoop ready. Do not use for ordinary feature development.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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. Distinctive:

    Only pause when an account login, browser confirmation, missing remote choice, or unsupported repository permission truly requires the user.

    line 14
  2. Very specific:

    If Codex authentication is missing or uses an API key, run codex login and direct the user to Sign in with ChatGPT.

    line 20
  3. Very specific:

    If GitHub CLI authentication is missing, run gh auth login --web, let the user complete the browser login, then rerun the initializer.

    line 21
  4. Very specific:

    If no GitHub origin exists, finish local setup, then ask for the intended OWNER/REPOSITORY and public/private choice.

    line 22
  5. Very specific:

    Verify local validation and tests again if any setup file changed.

    line 24
  6. Very specific:

    After READY, tell the user they can start with 作りたいもの: ...

    line 35

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
SKILL.md

Autoloop Setup

Make the current copied template ready for local work, GitHub Issues and pull requests, native Codex review, and fail-closed auto-merge. This setup uses the user's ChatGPT subscription login. Never request or configure OPENAI_API_KEY.

Authorization boundary

A direct setup request authorizes idempotent local initialization and repository-scoped GitHub settings for the current repository: labels, Actions variables, merge settings, and the supplied ruleset. It does not authorize creating a remote repository, choosing public visibility, connecting accounts, purchasing anything, or weakening a failed protection. --yolo changes CLI permissions but does not expand this scope.

Only pause when an account login, browser confirmation, missing remote choice, or unsupported repository permission truly requires the user. Complete every independent step before pausing.

Procedure

  1. Work from the repository root. Inspect git status, AGENTS.md, .autoloop/config.json, and the origin remote. Preserve unrelated user changes.
  2. Confirm Python 3.10+, Node.js 20+, Git, Codex CLI, and GitHub CLI are available. Run python scripts/autoloop/initialize_repository.py --apply-github. Use python3 only when python is unavailable.
  3. If Codex authentication is missing or uses an API key, run codex login and direct the user to Sign in with ChatGPT. Confirm with codex login status. Do not ask for an API key.
  4. If GitHub CLI authentication is missing, run gh auth login --web, let the user complete the browser login, then rerun the initializer.
  5. If no GitHub origin exists, finish local setup, then ask for the intended OWNER/REPOSITORY and public/private choice. Do not create or publish a repository without that choice.
  6. When the initializer succeeds, open https://chatgpt.com/codex/settings/code-review with the platform's normal browser opener. Ask the user to connect this repository and turn on both Code review and Automatic reviews if they are not already enabled. This account connection has no supported repository-side CLI setting, so do not claim it is enabled without confirmation.
  7. Verify local validation and tests again if any setup file changed. Verify AUTOLOOP_AUTO_MERGE=true, the active Codex AutoLoop main protection ruleset, and required checks policy-integrity, project-ci, and codex-review-gate.

Completion report

Report one of these outcomes clearly:

  • READY: local setup, ChatGPT login, GitHub settings, and the user's Codex review confirmation are complete.
  • LOCAL_READY: local setup is complete but GitHub origin or access is not configured.
  • ONE_TIME_ACTION: all automatable work is complete and the exact browser login or review toggle still needed is shown.
  • BLOCKED: a required protection cannot be applied; include the observed error and do not enable auto-merge.

After READY, tell the user they can start with 作りたいもの: ... or $autoloop ....

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.