minecode

MineCode is a local MCP server that gives AI assistants real-time access to Minecraft data, documentation, datapack generators, and your minecraft logs. It exposes a suite of tools (19 in total) across categories like Minecraft Wiki, Mojira, Spyglass, Misode Generators, and Logs, enabling tasks such as wiki searches, command documentation, bug tracking, registry and state inquiries, datapack generation, and log reading. To use it, install MineCode, start the server locally, and connect your MCP client (for example VS Code Copilot MCP or Claude Desktop) to the running minecode.server. The project provides a ready-to-use Python implementation that runs as a module, so you can launch it with Python’s -m flag and interact with the exposed tools through your MCP client.

Prerequisites:

• Python 3.10 or newer
• pip (comes with Python)

Install the MineCode MCP package from PyPI and set up a local environment:

# Create and activate a virtual environment (recommended)
python -m venv venv
# Windows
venv\Scripts\activate
# macOS/Linux
source venv/bin/activate

# Install MineCode MCP
pip install minecode-mcp

Run the MCP server locally:

python -m minecode.server

Configure your MCP client (e.g., VS Code Copilot MCP or Claude Desktop) to connect to the running server using the provided command. For VS Code Copilot, you typically add a configuration that points to the Python module minecode.server; for Claude Desktop, add a server entry under the mcpServers section with the appropriate command. You can also run the server in a virtual environment and verify the server starts listening on the expected interfaces.

Optional: to build or test locally, you can also install additional development dependencies as described in the repository's development notes, and you can run the release script if you’re publishing a new version.

Tips and notes:

• The server is designed for local development and integrates with various data sources (Minecraft Wiki, Mojira, Spyglass, Misode, and Logs). Ensure you have network access to these sources if using live querying.
• If you’re using Claude Desktop or another MCP client, ensure the client’s mcpServers configuration references the correct command to start the MineCode server (for Python, typically the module invocation as shown in the configuration examples).
• If you encounter connection issues, verify that the server is running (no port conflicts) and that your MCP client is pointed to the correct process (minecode.server).
• Virtual environments are recommended to isolate dependencies. When using a venv, remember to activate it before launching the server.
• For development and packaging, follow the repository’s guidance on release scripts and token handling for PyPI publishing if you contribute builds.