Skip to the collection

agent-native · Agent Runtime & Infrastructure

Codex plugin for Claude Code (codex-plugin-cc)

Listed on The Agent-Native Index

Use Codex from inside Claude Code for code reviews or to delegate tasks to Codex

What it does

The Codex plugin for Claude Code lets a Claude Code session invoke OpenAI Codex for read-only reviews, adversarial design review, and delegated rescue work (investigate, patch, resume threads) without leaving the Claude Code workflow. It exposes slash commands (/codex:review, /codex:adversarial-review, /codex:rescue, /codex:status, /codex:result, /codex:cancel, /codex:setup) and registers a Codex subagent for handoffs.

Background jobs integrate with status/result/cancel so an orchestrating agent can run long Codex tasks asynchronously. An optional review gate (/codex:setup --enable-review-gate) hooks Claude’s stop flow to a targeted Codex review before completion.


Agent-facing interfaces

How agents onboard

Interaction pattern: Extension / Plugin (Claude Code marketplace plugin wrapping local Codex CLI + app server)

# In Claude Code: add marketplace and install plugin (see repo README)
/plugin marketplace add openai/codex-plugin-cc
/plugin install codex@openai-codex
/reload-plugins
/codex:setup

# Optional: ensure Codex CLI is installed and signed in
npm install -g @openai/codex
# then from Claude Code: !codex login

# Typical delegation flow
/codex:review --background
/codex:status
/codex:result

# Delegate a fix or investigation to Codex (subagent)
/codex:rescue investigate why the tests started failing

See the full onboarding section in the dossier.

Protocol surface

Interface Detail
Claude Code plugin Marketplace: openai/codex-plugin-cc — package codex@openai-codex
Slash commands /codex:* as documented in repo README
Codex CLI Global codex binary — config.toml / .codex/config.toml per Codex config
Codex app server Used by the plugin for Codex integration — app server docs