Get the FREE Ultimate OpenClaw Setup Guide →

robot

mcp server for robot and automations

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio showkeyjar-robot-mcp-server python -m robot_mcp_server \
  --env LOG_LEVEL="INFO" \
  --env PYTHONUNBUFFERED="1" \
  --env ROBOT_SDK_VERBOSE="0"

How to use

This MCP server implements a Robot control interface following the Model Context Protocol (MCP). It exposes two main tool groups:宇树机器人 tools and 大疆无人机 (DJI) tools. The宇树工具 allow you to establish a connection to the Unitree robot, move it with a specified velocity, and stop immediately if needed. TheDJI tools enable establishing a connection to a DJI Tello drone and commanding takeoff to a given height or performing a safe landing. To use it, connect a MCP client (for example via the MCP Python client) and call the provided tools by name, passing the expected parameters.

Typical workflow:

  • Connect the MCP client to the server with Client.connect_stdio() or the appropriate transport.
  • Call unitree_connect to initialize communication with the Unitree robot, followed by unitree_move with a velocity and duration (or implement continued movement as needed).
  • Call dji_connect to initialize the DJI drone, then dji_takeoff with a height parameter to rise to a target altitude, and dji_land to land safely when finished.

The server adheres to the MCP standard interfaces, supports real-time status monitoring, and includes an emergency stop mechanism for both robot and drone operations. Use the logging facilities to diagnose issues and monitor system state during development and operation.

How to install

Prerequisites:\n- Python 3.10+\n- Virtual environment support (recommended)\n- Access to install Python packages from the internet (pip)\n\nInstallation steps:\n1) Create and activate a virtual environment:\n python -m venv .venv\n # Windows\n .venv\Scripts\activate\n # macOS/Linux\n source .venv/bin/activate\n\n2) Install robot and drone SDKs (as described in the project):\n pip install git+https://github.com/unitreerobotics/unitree_sdk2_python.git\n pip install djitellopy\n\n3) Install other dependencies (if a requirements.txt is provided):\n pip install -r requirements.txt\n\n4) Run the MCP server (see mcp_config for details):\n python -m robot_mcp_server\n\n5) Verify the server is running by connecting a MCP client and issuing tool calls as documented in the README.

Additional notes

Tips and notes:\n- Ensure the robot hardware or simulators are connected as required before issuing control commands.\n- The installation steps auto-install the required SDKs (Unitree and DJI Tello) as indicated in the README. If automatic installation fails, install the SDKs manually according to their respective docs.\n- Use the emergency stop tools to halt motion immediately in case of errors.\n- Check logs for operation statuses, connection states, and error messages to troubleshoot integration issues.\n- If deploying in a containerized environment, expose the appropriate ports and mount device access to the container for robot/drone SDKs to function correctly.

Related MCP Servers

Sponsor this space

Reach thousands of developers