Codexclaw
Telegram bot for Claude Code and Codex CLI with MCP routing, multi-agent orchestration, cron jobs, and access controls.
What is Codexclaw?
Codexclaw is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to telegram bot for claude code and codex cli with mcp routing, multi-agent orchestration, cron jobs, and access controls.
Telegram bot for Claude Code and Codex CLI with MCP routing, multi-agent orchestration, cron jobs, and access controls.
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Telegram bot for Claude Code and Codex CLI with MCP routing,
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx codexclawConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Codexclaw
CodexClaw is a Telegram bot that brings Claude Code and OpenAI Codex CLI capabilities directly into Telegram chats, with built-in MCP routing, multi-agent orchestration, cron-based automation, and per-user access controls. It supports two execution backends — the OpenAI Codex SDK (preferred) or a legacy CLI/PTY mode — and maintains per-chat conversation state across project switches. Developers use it to delegate coding tasks, manage GitHub repos, and run automated workflows entirely from Telegram without opening a terminal.
Prerequisites
- Node.js 18 or later installed
- A Telegram bot token from @BotFather
- OpenAI API key (for Codex SDK backend) or Codex CLI installed
- Git and a defined WORKSPACE_ROOT directory containing your projects
- An MCP-compatible client or the CodexClaw bot itself as the interface
Clone and install dependencies
Clone the CodexClaw repository and install Node.js dependencies.
git clone https://github.com/MackDing/CodexClaw.git
cd CodexClaw
npm installCreate your environment file
Copy the example environment file and fill in your credentials and paths.
cp .env.example .envConfigure required environment variables
Edit the .env file with your Telegram bot token, allowed user IDs, workspace settings, and backend selection. BOT_TOKEN comes from @BotFather; ALLOWED_USER_IDS is a comma-separated whitelist of Telegram user IDs.
BOT_TOKEN=your_telegram_bot_token
ALLOWED_USER_IDS=123456789,987654321
STATE_FILE=/path/to/state.json
WORKSPACE_ROOT=/home/user/projects
CODEX_WORKDIR=/home/user/projects/myproject
CODEX_BACKEND=sdk
CODEX_SDK_SANDBOX_MODE=workspace-write
CODEX_SDK_APPROVAL_POLICY=neverStart the bot
Launch the CodexClaw bot. It will connect to Telegram and begin listening for commands from authorized users.
npm run startTest bot connectivity
Open Telegram, find your bot by the username you set with @BotFather, and send /status to confirm the bot is running and shows the correct configuration.
Switch projects and run tasks
Use /repo to list and switch between Git projects in your WORKSPACE_ROOT, then use /exec or /auto to delegate coding tasks.
Codexclaw Examples
Client configuration
CodexClaw runs as a standalone Telegram bot rather than a traditional MCP server config. Below is the .env configuration for SDK backend mode.
{
"BOT_TOKEN": "your_telegram_bot_token_from_botfather",
"ALLOWED_USER_IDS": "123456789",
"STATE_FILE": "/home/user/.codexclaw/state.json",
"WORKSPACE_ROOT": "/home/user/projects",
"CODEX_WORKDIR": "/home/user/projects/myapp",
"CODEX_BACKEND": "sdk",
"CODEX_SDK_SANDBOX_MODE": "workspace-write",
"CODEX_SDK_APPROVAL_POLICY": "never",
"GITHUB_TOKEN": "ghp_your_github_token",
"CRON_DAILY_SUMMARY": "0 9 * * *"
}Prompts to try
Telegram commands and task descriptions to use with CodexClaw once the bot is running.
- "/status" — Show current configuration and session state
- "/repo" — List available Git projects and switch between them
- "/exec Refactor the authentication module to use JWT tokens"
- "/auto Write unit tests for the payment service"
- "/gh commit 'Add rate limiting middleware'"
- "/mcp list" — Show available MCP servers and their status
- "/plan Build a REST API endpoint for user profile updates"Troubleshooting Codexclaw
Bot does not respond to messages
Check that your Telegram user ID is listed in ALLOWED_USER_IDS. Obtain your Telegram user ID by messaging @userinfobot. Also verify BOT_TOKEN is correct and the bot is running without errors in the console.
Codex execution fails or hangs
If using CODEX_BACKEND=sdk, ensure your OpenAI API key is set and has access to the Codex model. If using cli backend, verify the codex binary is installed and CODEX_COMMAND points to its path. Check CODEX_WORKDIR exists and is accessible.
Contention error when multiple chats try to write to the same project
CodexClaw blocks simultaneous writes to the same CODEX_WORKDIR from different chats. Wait for the active task to complete, or use /new in the other chat to reset its session state.
Frequently Asked Questions about Codexclaw
What is Codexclaw?
Codexclaw is a Model Context Protocol (MCP) server that telegram bot for claude code and codex cli with mcp routing, multi-agent orchestration, cron jobs, and access controls. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Codexclaw?
Follow the installation instructions on the Codexclaw GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Codexclaw?
Codexclaw works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Codexclaw free to use?
Yes, Codexclaw is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Codexclaw Alternatives — Similar Coding Agents Servers
Looking for alternatives to Codexclaw? 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 Codexclaw 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 Codexclaw?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.