emacs

This MCP server enables generating Emacs Lisp code tailored to a task and then executing that code in a running Emacs instance. It exposes two tools: generate_emacs_lisp_code, which produces elisp snippets for a given objective, and execute_emacs_lisp_code, which sends those snippets to Emacs via emacsclient for evaluation. Use the generate tool to obtain code that accomplishes a user-specified task, then pass that code to the execute tool so it runs inside a live Emacs session. The workflow is designed to integrate seamlessly with Claude Desktop, letting you prototype and run Emacs automations from chat.

Prerequisites:

• Python 3.8+ and a working Python virtual environment
• The uv (uvicorn-like) tooling required by the project
• Access to the Emacsclient binary for remote evaluation

Install and run steps:

1. Install dependencies and set up the environment

uv sync
source .venv/bin/activate

2. Run the MCP server in development mode

mcp dev emacs_mcp_server.py

3. Install the MCP server so Claude Desktop can discover it

mcp install emacs_mcp_server.py

Note: After installation, edit claude_desktop_config.json to provide API keys and the path to emacsclient as shown in the example configuration.

Environment variables you may need:

• OPENAI_API_KEY: Your OpenAI API key for model access
• EMACSCLIENT: Absolute path to your emacsclient executable (required for remote evaluation)

Configuration tips:

• Ensure Emacs is running with emacsclient enabled (server mode) if you plan to eval code in an existing session.
• The mcp_config here uses uv to run the server; you can adapt the command path if you deploy differently.
• When testing, start with small elisp snippets to confirm communication between the MCP server and Emacs. If emacsclient cannot connect, verify networking, permissions, and the EMACSCLIENT path.