AIAny
AI Client2026
Icon for item

Codex plugin for Claude Code

Integrates Codex into Claude Code so you can run read-only code reviews, steerable adversarial reviews, and delegate long-running tasks to a local Codex instance via slash commands. Uses the local Codex CLI/app server and Node.js; designed for developers who want seamless handoff between Claude Code and Codex.

Introduction

The plugin fills a pragmatic gap: developers often start debugging or design conversations in Claude Code but need Codex for longer-running, session-based, or model-specific work. By wiring Claude Code to your local Codex CLI and app server, the plugin makes it easy to hand off review and remediation tasks without ripping context out of your interactive session. This reduces context-switch overhead and lets you run steerable, background, or larger-model tasks where appropriate.

What Sets It Apart
  • Direct handoff from Claude Code to a local Codex runtime: you can transfer the current session or delegate work so the Codex thread preserves conversation history and repository context, so follow-up work continues with the same transcript.
  • Multiple purpose-built slash commands: /codex:review (read-only code review), /codex:adversarial-review (steerable challenge reviews), /codex:rescue (delegate bug investigation or fixes), plus /codex:transfer, /codex:status, /codex:result, /codex:cancel. This maps common developer workflows to single commands rather than manual exports.
  • Background and resume-friendly workflows: long-running or resource-heavy tasks can be forced into background, produce session IDs, and be resumed directly inside Codex; config and model defaults follow your local Codex settings, so project-level preferences carry over.
  • Local-first security and config: the plugin delegates through the Codex CLI/app server on your machine, so it uses local authentication, local repo checkout, and the same environment—helpful for reproducibility and for using private credentials, but also a trust boundary to consider.
Who It's For and Tradeoffs

Great fit if you use Claude Code for interactive exploration but want Codex for longer, model-specific, or resumable work; if you already run Codex locally and want quick, in-session delegation without manual exporting. Look elsewhere if you need a pure cloud-hosted handoff (the plugin expects a local Codex runtime), or if you expect the plugin to autonomously change code during reviews (review commands are read-only; fixes must be applied via delegated rescue tasks or in Codex itself). Also note the optional review-gate can create loops and consume usage limits if misconfigured.

Information

  • Websitegithub.com
  • OrganizationsOpenAI
  • Authorsdkundel-openai, VOIDXAI, zeta987, xiaolai, Bhuvanesh09, pengyou200902, rgbkrk, manavagarwal123, omid-web, Pranav322
  • Published date2026/03/30

More Items

GitHub
AI Client2026

Runs research, writing, code review, and cross-app tasks from your desktop, with controls over what its agents can do. It works with cloud or local models and pairs approvals with activity logs and sandboxing.

GitHub
AI Agent2026

Orchestrates coding agents through repeatable software delivery workflows, with sandboxed workspaces, human approval gates, and run history. Teams can use chat for exploratory work or standardize proven processes as SOPs.

GitHub
AI Coding2026

Switches Claude Code between saved accounts manually or as usage limits approach, with per-account quota tracking and isolated parallel sessions. It manages local credentials rather than changing any account’s subscription limits.