ros2_mcp
An MCP server that enables MCP clients to interact with ROS 2 systems, providing runtime access and safe project creation and modification.
README
ROS 2 MCP
A modular MCP server for connecting MCP clients such as Codex to ROS 2.
Goals
The project provides two clearly separated capabilities:
- Runtime access to a running ROS 2 system
- Safe creation and modification of ROS 2 projects
The architecture keeps MCP, application logic, ROS integration, and project management separated.
Architecture
MCP Client
|
v
MCP Layer
|
v
Application Layer
/ \
v v
ROS Project
Adapter Adapter
| |
v v
rclpy Safe Filesystem
|
v
ROS 2 / DDS
Project Principles
- Modular architecture
- MCP and ROS 2 remain separated
- ROS access goes through a dedicated adapter
- ROS distributions should remain replaceable
- MCP should remain as stateless as possible
- MCP clients should remain replaceable
- Runtime access starts read-only
- Write operations require explicit safety mechanisms
- Project file access is restricted to approved project roots
- No unnecessary frameworks
- Docker and Kubernetes are planned for later phases
Development Phases
- Foundation
- ROS Adapter
- ROS Discovery
- Topic Reading
- Project Adapter
- Build and Test
- Extended Runtime
- Controlled Write and Deployment Readiness
Detailed documentation for each phase is stored in docs/README_PHASE_X.md.
Current Status
Phase 7 completed. Phase 8 is next.
Development Environment
- Ubuntu 24.04
- ROS 2 Jazzy
- Python 3.12
- uv
- MCP Python SDK
- pytest
Repository
This repository is developed independently and is intended to be published later as its own GitHub repository.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。