Codex plugin for Claude Code
OpenAI's official plugin: Codex reviews of your changes and task delegation to Codex from inside a Claude Code session
Install
/plugin marketplace add openai/codex-plugin-cc
/plugin install codex@openai-codex
/reload-plugins
/codex:setup/codex:setup checks whether Codex is installed and can offer to install it via npm.
This is third-party code. Review the repository files before installing.
What it does
The plugin adds Claude Code commands that call your locally installed Codex CLI. /codex:review runs a standard review of uncommitted changes or a branch against a base, and /codex:adversarial-review challenges design decisions around a focus you give it. /codex:rescue hands a task to a Codex subagent, and /codex:transfer moves the current session into Codex so you can continue there. Background jobs are tracked with /codex:status, /codex:result and /codex:cancel, and an optional review gate runs on the Stop hook.
Who it is for. For Claude Code developers who want a second model's opinion or a backup agent.
Good fit when
- You want an independent review before merging
- You want to stress-test a design choice: races, rollback, auth
- Claude is stuck on a bug and the task should go to Codex
Not a fit when
- You have neither a ChatGPT account nor an OpenAI API key
- You need a review without sending code to an external provider
Example request
Ask Codex to review my branch against main in the backgroundLimitations
Requires Node.js 18.18 or later, the Codex CLI and sign-in with a ChatGPT account or OpenAI API key; usage counts toward Codex limits. The review gate can create a long Claude and Codex loop that drains limits quickly. OpenAI and Anthropic services are not available from Russia without a VPN.
How to disable. Turn off the review gate with /codex:setup --disable-review-gate, then uninstall the codex plugin via /plugin.
Security check
- Sends code and session context to OpenAI
- Rescue tasks can modify project files
README in short
The README lists the plugin's commands, requirements and install steps through the openai/codex-plugin-cc marketplace. Each command is documented with its flags, such as --base, --background, --wait, --resume and --fresh, plus typical flows. The plugin wraps the Codex app server and uses the same binary, sign-in and config as regular Codex. It warns that the review gate can cause long loops and drain limits.
FAQ
Do I need a separate Codex account?
No, the plugin uses the Codex CLI's local sign-in. If you have not signed in yet, a ChatGPT account or API key works.
Does it use my existing Codex config?
Yes, it reads the same ~/.codex/config.toml and the project .codex/config.toml when the project is trusted.
Related
A skills library that gives coding agents a development process: brainstorming, planning, TDD, subagents and code review
Skills for real engineers by Matt Pocock
Skills For Real Engineers
Small composable skills for engineering with agents: plan grilling, TDD, bug diagnosis, code review and architecture
Reference MCP servers
Model Context Protocol servers
Official reference MCP servers: Filesystem, Fetch, Git, Memory, Sequential Thinking, Time and Everything
Up-to-date, version-specific library docs and code examples in your agent's context, via MCP or a CLI plus skill