ros-mcp-server and robotmcp_client
The ros-mcp-server is the MCP server implementation that exposes ROS functionality, while robotmcp_client is the client library for connecting to that server, making them complements designed to be used together.
About ros-mcp-server
robotmcp/ros-mcp-server
Connect AI models like Claude & GPT with robots using MCP and ROS.
Implements MCP (Model Context Protocol) as a ROS/ROS2 bridge, enabling LLMs to introspect and control robot systems through topic publishing, service calls, parameter management, and real-time sensor subscriptions without modifying existing robot code. Supports both ROS1 and ROS2 ecosystems through a single deployment, integrating with MCP-compatible clients (Claude Desktop, Gemini, ChatGPT) via rosbridge.
About robotmcp_client
robotmcp/robotmcp_client
Connect AI models like Claude & GPT with robots using MCP and ROS.
Implements the MCP client specification to enable direct LLM-robot communication over stdio or HTTP transports, supporting bidirectional streaming of commands and sensor feedback. Includes a production-ready Gemini Live client with real-time interaction capabilities, plus abstractions for integrating alternative LLMs (Ollama, Nvidia NeMo). Works offline within local/LAN environments and is version-agnostic across ROS1 and ROS2 via rosbridge integration.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work