codex-mcp-tool
by x51xxx·★ 21·Score 43
MCP server connecting Claude/Cursor to Codex CLI with file analysis, multi-turn conversations, and local model support.
Overview
Codex MCP Server provides a bridge between Claude/Cursor and the Codex CLI, enabling advanced code analysis through file references (@src/, @package.json), multi-turn conversations with session management, and structured change modes for refactoring. The server supports various models including GPT-5.5 and local OSS models via Ollama or LM Studio. It features sandboxed code execution, web search capabilities, and brainstorming frameworks like SCAMPER.
Try asking AI
After installing, here are 5 things you can ask your AI assistant:
When to choose this
Choose this server when you're already using Codex CLI and want seamless integration with Claude/Cursor for code analysis, especially for multi-turn conversations about your codebase.
When NOT to choose this
Avoid if you need write access to files (this is read-only) or if you're not already using Codex CLI, as the setup requires additional CLI installation beyond MCP configuration.
Tools this server exposes
5 tools extracted from the READMEask-codexExecute Codex CLI with file analysis, models, sessions
brainstormGenerate ideas with SCAMPER, design-thinking, etc.
list-sessionsView/delete/clear conversation sessions
healthDiagnose CLI installation, version, features
pingTest connection with Codex CLI
Comparable tools
Installation
Installation
Claude Desktop
claude mcp add codex-cli -- npx -y @trishchuk/codex-mcp-toolConfiguration
{
"mcpServers": {
"codex-cli": {
"command": "npx",
"args": ["-y", "@trishchuk/codex-mcp-tool"]
}
}
}**Config locations:**
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
FAQ
- What do I need to install before using this MCP server?
- You need Node.js 18+, the Codex CLI installed, and authentication with OpenAI configured. The server works with Claude Desktop.
- Can I use local models with this MCP server?
- Yes, v1.6.0+ supports local OSS models via Ollama or LM Studio. You need to have Ollama running with a model that supports tool calling.
Compare codex-mcp-tool with
Last updated · Auto-generated from public README + GitHub signals.