$ loading_
Connect Claude Code and OpenAI Codex CLI for plan-execute-review coding workflows.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "codex-mcp" yet — see the docs or source repo.
First ask me about the goals, constraints, and risks of this legacy module, then design a refactoring plan; after confirmation, have Codex implement the code changes and finally review the diff for potential issues.
You get requirement clarification and a refactoring plan first, followed by code changes and a review with risks noted.
Interview me about the requirements first, break down the implementation steps and edge cases, then have Codex build the feature and review whether the implementation matches the design.
It produces clarified requirements, an implementation plan, actual code changes, and a final review of the result.
I will provide the error symptoms and relevant code. First analyze likely causes and confirm the fix direction with me, then have Codex modify the code, and finally review whether the fix is complete and note any remaining risks.
You receive issue analysis, a fix plan, code changes, and a post-fix review summary.
When handling complex changes, developers can use it to let Claude handle interviewing, planning, and review while Codex performs the implementation. This fits workflows that need discussion before execution.
When a team wants to clarify requirements and constraints before making changes, this tool supports an interactive plan-execute-review flow. It helps reduce mismatches from coding too early.
If you want an additional review step after code generation, it lets Claude inspect whether Codex’s implementation matches the plan and identify possible issues. This is useful in quality-sensitive development work.
It is an MCP tool that connects Claude Code with the OpenAI Codex CLI. Its core workflow lets Claude handle interviewing, design, and review while Codex implements code changes.
The description emphasizes a role-based workflow rather than a single assistant doing everything at once. It separates planning, execution, and review across Claude and Codex.
From the name and description, it depends on Claude Code, the OpenAI Codex CLI, and an MCP environment. The provided materials do not include exact installation or configuration steps; see the source repository.
Call Codex inside Claude Code for reviews, second opinions, and delegated coding.
Coordinate Codex and Claude Code through a shared task board and inbox.
Connect OpenAI Codex CLI to MCP clients for headless coding sessions.
Orchestrate a local Claude Code and Codex CLI development workflow.
Use Codex in Claude Code for execution, debugging, and web search.
Enable Claude Code to perform coding tasks through the OpenAI Codex CLI.