ROS2 MCP Server

A Model Context Protocol (MCP) server that provides AI assistants with tools to introspect and interact with ROS2 robotics systems. This server enables Large Language Models (LLMs) and other AI systems to understand, monitor, and analyze ROS2 robot systems in real-time.

Overview

The ROS2 MCP Server bridges the gap between AI assistants and robotics systems by exposing ROS2 functionality through the Model Context Protocol. It provides tools for topic monitoring, sensor data collection, system introspection, and real-time robot state analysis.

Features

Core Capabilities

• Topic Monitoring: Echo messages from any ROS2 topic with configurable message counts and timeouts
• Image Processing: Direct image retrieval from ROS2 Image topics for VLM (Vision Language Model) analysis
• System Introspection: List all available topics, nodes, and message types
• Interface Analysis: Detailed inspection of ROS2 message/service/action interfaces
• Real-time Status: Live system health monitoring and status reporting

AI-Friendly Design

• Structured Prompts: Pre-built prompts for common robotics analysis tasks
• Resource Streaming: Live system status via MCP resources
• VLM Integration: Direct image data provision for computer vision analysis
• Error Handling: Robust error reporting and fallback mechanisms

Installation

Prerequisites

• ROS2 (Humble, Iron, or Jazzy)
• Python 3.8+
• OpenCV (cv_bridge package)

using in your ROS2 workspace

To use this module in your ROS2 workspace, follow these steps:

1. Clone the repository into your ROS2 workspace’s src directory:

   cd ~/your_ros2_workspace/src
   git clone 
   

2. Install dependencies using rosdep to automatically handle both ROS and pip dependencies:

   cd ~/your_ros2_workspace
   rosdep install -i --from-paths src/
   

   This will install the required mcp[cli] pip dependency along with any other ROS dependencies.

3. Build the workspace:

   colcon build
   

4. Source the workspace:

   source install/setup.bash
   

5. Run the nodes as needed:

   ros2 run ros2_mcp 
   

Note: Make sure you have Python 3 and pip installed on your system, as the MCP CLI tools require Python dependencies that will be installed automatically through rosdep.

Docker Usage

Pre-built Images

Docker images are automatically built and published to the L-CAS registry:

docker pull lcas.lincoln.ac.uk/:latest

Building Locally

To build the Docker image locally:

docker build -t ros2-mcp .

Running the Container

# Run with ROS2 network access
docker run -it --rm --network host ros2-mcp

# Run with custom ROS_DOMAIN_ID
docker run -it --rm --network host -e ROS_DOMAIN_ID=42 ros2-mcp

Docker Compose (Recommended)