Claude Codex
Claude Code workflow integrating OpenAI Codex as an MCP server — plan, implement, and review loops with smart routing between Claude and Codex.
What is Claude Codex?
Claude Codex is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to claude code workflow integrating openai codex as an mcp server — plan, implement, and review loops with smart routing between claude and codex.
Claude Code workflow integrating OpenAI Codex as an MCP server — plan, implement, and review loops with smart routing between Claude and Codex.
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Claude Code workflow integrating OpenAI Codex as an MCP serv
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx claude-codexConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Claude Codex
Claude Codex is an MCP server that integrates OpenAI Codex as a code review and implementation backend within Claude Code workflows, enabling intelligent routing between Claude and Codex based on task size and type. It exposes five slash-command-driven skills that implement plan-audit, implement-review, TDD, and smart-routing loops — Claude plans or implements while Codex audits for correctness, completeness, and security with structured APPROVED/WARNING/BLOCKED verdicts. Development teams use it when they want the reasoning strength of Claude combined with Codex's code-specific review capabilities in an automated feedback loop.
Prerequisites
- Claude Code CLI installed (claude command available in PATH)
- OpenAI Codex CLI installed globally: npm install -g @openai/codex
- An OpenAI API key with access to Codex models (set as OPENAI_API_KEY environment variable)
- Node.js 18+ for running the Codex MCP server
- Git to clone the repository
Install OpenAI Codex CLI globally
The claude-codex integration requires the @openai/codex package to be installed globally so it can be launched as an MCP server subprocess.
npm install -g @openai/codexClone the claude-codex repository
Check out the project which contains the skills, prompts, and commands that define the integration workflows.
git clone https://github.com/ching-kuo/claude-codex.git
cd claude-codexCopy skills, prompts, and commands to Claude Code config
Install the workflow definitions into your Claude Code user configuration directory so the slash commands become available in your Claude Code sessions.
cp -r skills/* ~/.claude/skills/
cp -r prompts/* ~/.claude/prompts/
cp -r commands/* ~/.claude/commands/Set your OpenAI API key
Export your OpenAI API key in your shell profile or current session so Codex can authenticate when launched as an MCP server.
export OPENAI_API_KEY=sk-your-openai-api-key-hereRegister the Codex MCP server with Claude Code
Use the claude mcp add command to register Codex as an MCP server at the user scope. This configures Codex to run with high reasoning effort and the gpt-5.3-codex model.
claude mcp add codex -s user -- codex -c model=gpt-5.3-codex -c model_reasoning_effort=high mcp-serverVerify the integration in a project
Open a terminal in your project directory and start Claude Code. Verify the Codex MCP server is connected and try a planning command.
claude
# Then inside claude:
/plan-codex Add JWT authentication to the APIClaude Codex Examples
Client configuration
The Codex MCP server is registered via the Claude Code CLI rather than a JSON config file. This is the equivalent configuration that claude mcp add creates in your Claude Code user settings.
{
"mcpServers": {
"codex": {
"command": "codex",
"args": ["-c", "model=gpt-5.3-codex", "-c", "model_reasoning_effort=high", "mcp-server"],
"env": {
"OPENAI_API_KEY": "sk-your-openai-api-key-here"
}
}
}
}Prompts to try
Claude Codex exposes five slash commands in Claude Code. Use these workflows for plan-audit and implement-review loops.
- /plan-codex Add rate limiting middleware to the Express API
- /claude-codex Add input validation to the user registration endpoint
- /execute-codex .claude/plan/rate-limiting.md
- /tdd-claude-codex Implement a binary search function with full test coverage
- /tdd-execute-codex .claude/plan/search-feature.mdTroubleshooting Claude Codex
codex command not found when Claude Code tries to start the MCP server
Ensure @openai/codex is installed globally with npm install -g @openai/codex and that your global npm bin directory is in your PATH. Run which codex to verify. If npm global bin is not in PATH, add it: export PATH="$(npm bin -g):$PATH".
Codex returns authentication errors when invoked from within Claude Code
The OPENAI_API_KEY environment variable must be available in the environment where Claude Code is launched, not just your current terminal session. Add it to your shell profile (~/.zshrc or ~/.bashrc) so it is always exported, then restart Claude Code.
Slash commands /plan-codex or /claude-codex are not available in Claude Code
Verify the files were copied to the correct directories: ~/.claude/skills/, ~/.claude/prompts/, and ~/.claude/commands/. Claude Code loads these at startup — restart Claude Code after copying. Run ls ~/.claude/commands/ to confirm the command files are present.
Frequently Asked Questions about Claude Codex
What is Claude Codex?
Claude Codex is a Model Context Protocol (MCP) server that claude code workflow integrating openai codex as an mcp server — plan, implement, and review loops with smart routing between claude and codex. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Claude Codex?
Follow the installation instructions on the Claude Codex GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Claude Codex?
Claude Codex works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Claude Codex free to use?
Yes, Claude Codex is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Claude Codex Alternatives — Similar Coding Agents Servers
Looking for alternatives to Claude Codex? Here are other popular coding agents servers you can use with Claude, Cursor, and VS Code.
Dify
★ 142.2kProduction-ready platform for agentic workflow development.
Ruflo
★ 54.0k🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integrat
Goose
★ 45.7kan open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
Antigravity Awesome Skills
★ 38.3kInstallable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
AgentScope
★ 25.5kBuild and run agents you can see, understand and trust.
Serena
★ 24.5kA coding agent toolkit that provides IDE-like semantic code retrieval and editing tools, enabling LLMs to efficiently navigate and modify codebases using symbol-level operations instead of basic file reading and string replacements.
Browse More Coding Agents MCP Servers
Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Claude Codex in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Claude Codex?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.