Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Republish for discoverability check.
What This Skill Does
Runs OpenAI Codex CLI commands from OpenClaw against a target project directory. Supports one-shot tasks, interactive sessions, and long-running background jobs with process tracking and input relay.
Wraps Codex CLI with PTY and workdir configuration so interactive or long-running coding tasks don't hang or lose context inside OpenClaw.
When to use it
- Implementing a new feature with tests in one command
- Finding the root cause of a failing CI build
- Reviewing a branch diff for high-risk changes before merging
- Refactoring a module across an unfamiliar codebase
- Running scripted multi-step coding workflows in the background
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: implement a rate-limiting middleware with tests
- 1Verify Codex CLI is available by running `codex --version`
- 2Start a background exec with `codex exec "Implement rate-limiting middleware with tests, run tests, summarize changed files"` pointed at the target repo
- 3Record the session ID and poll the process for completion milestones
- 4Read output logs to confirm test results and which files were modified
- 5Report the implementation summary and test outcomes to the user
Rate-limiting middleware implemented, tests passing, list of changed files returned
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.