io.github.ReGenNow/regennexus-mcp
编码与调试by regennow
面向 IoT 的 AI 控制硬件 MCP,支持 GPIO、robotics、sensors、cameras 与 mesh networking 等能力。
什么是 io.github.ReGenNow/regennexus-mcp?
面向 IoT 的 AI 控制硬件 MCP,支持 GPIO、robotics、sensors、cameras 与 mesh networking 等能力。
README
RegenNexus MCP Server
<img src="logo.png" alt="RegenNexus Logo" width="64" height="64"> <!-- mcp-name: io.github.ReGenNow/regennexus-mcp -->MCP (Model Context Protocol) server adapter for RegenNexus UAP - enabling AI-controlled hardware.
This package exposes RegenNexus hardware capabilities as MCP tools, allowing Claude Code, Claude Desktop, and other MCP-compatible AI clients to control physical devices.
Features
- Hardware Control: GPIO, PWM, robotic arms, sensors, cameras
- Mesh Networking: Discover and communicate with nodes over LAN (UDP/WebSocket)
- Serial/I2C: Communicate with microcontrollers and sensors
- Two Connection Modes: Local (direct import) or Remote (HTTP API)
- Auto-Discovery: Automatically detects installed RegenNexus and mesh nodes
- MCP Compatible: Works with Claude, Codex, Gemini, and any MCP client
Installation
# MCP server only (for remote UAP connection)
pip install regennexus-mcp
# With local UAP support
pip install regennexus-mcp[local]
# With remote API support
pip install regennexus-mcp[remote]
# Everything
pip install regennexus-mcp[all]
Quick Start
1. Configure Claude Code
Add to your Claude Code MCP settings:
{
"mcpServers": {
"regennexus": {
"command": "regennexus-mcp"
}
}
}
2. Use with Claude
Once configured, Claude can control hardware:
User: "List all connected devices"
Claude: [calls list_devices tool]
User: "Turn on GPIO pin 18"
Claude: [calls gpio_write with pin=18, value=1]
User: "Move the robot arm to position [0, 45, -30, 0, 60, 0, 0]"
Claude: [calls robot_arm_move with positions]
Configuration
Configure via environment variables:
# Connection mode: auto, local, or remote
REGENNEXUS_MODE=auto
# Remote mode settings
REGENNEXUS_ENDPOINT=https://your-uap-server.com
REGENNEXUS_API_KEY=your-api-key
# Local mode settings
REGENNEXUS_CONFIG=/path/to/regennexus-config.yaml
# Logging
REGENNEXUS_LOG_LEVEL=INFO
Claude Code Config with Environment
{
"mcpServers": {
"regennexus": {
"command": "regennexus-mcp",
"env": {
"REGENNEXUS_MODE": "local",
"REGENNEXUS_LOG_LEVEL": "DEBUG"
}
}
}
}
Available Tools (Free Tier)
These tools are provided by the regennexus package (open source):
GPIO & Basic I/O
| Tool | Description |
|---|---|
gpio_write | Set a GPIO pin to HIGH (1) or LOW (0) |
gpio_read | Read the current state of a GPIO pin |
pwm_write | Set PWM duty cycle (0-100%) for motors, LEDs, servos |
Sensors & I2C
| Tool | Description |
|---|---|
read_sensor | Read value from a sensor (temperature, humidity, etc.) |
i2c_scan | Scan I2C bus for connected devices |
Serial Communication
| Tool | Description |
|---|---|
serial_send | Send data over serial port (UART) |
serial_read | Read data from serial port |
Robotics
| Tool | Description |
|---|---|
robot_arm_move | Move a robotic arm to specified joint positions |
gripper_control | Open or close a robotic gripper |
Device Management
| Tool | Description |
|---|---|
list_devices | List all connected hardware devices |
device_info | Get device details (CPU, memory, IP, temperature) |
Camera
| Tool | Description |
|---|---|
camera_capture | Capture a single image from a camera |
Mesh Network
| Tool | Description |
|---|---|
list_nodes | List all nodes in the mesh network |
ping_node | Ping a node and measure network latency |
send_to_node | Send a message/command to a specific node |
broadcast_message | Broadcast a message to all nodes |
find_by_capability | Find nodes with a specific capability |
Premium tools with additional capabilities are available separately.
Connection Modes
Local Mode (Recommended)
Requires regennexus package installed on the same machine. Directly imports UAP modules for minimal latency. No API server needed.
pip install regennexus-mcp[local]
REGENNEXUS_MODE=local regennexus-mcp
This is the recommended mode for most users.
Remote Mode (Advanced)
Connects to a UAP instance running as a service via HTTP API. Requires UAP API server to be running (regen server).
# On the UAP server machine:
regen server --port 8080
# On the Claude Code machine:
pip install regennexus-mcp[remote]
REGENNEXUS_MODE=remote \
REGENNEXUS_ENDPOINT=http://uap-server:8080 \
regennexus-mcp
Auto Mode (Default)
Tries local first, falls back to remote if UAP not installed.
regennexus-mcp # Auto-detects best mode
Running Standalone
# Run MCP server
regennexus-mcp
# Or via Python
python -m regennexus_mcp
Architecture
Claude Code (MCP Client)
│
▼
regennexus-mcp (This Package)
│
├── Local Mode: Direct import
│ │
│ ▼
│ regennexus (UAP)
│
└── Remote Mode: HTTP API
│
▼
UAP Service
Development
# Clone
git clone https://github.com/regennow/regennexus-mcp
cd regennexus-mcp
# Install dev dependencies
pip install -e ".[dev]"
# Run tests
pytest
# Format code
black src/
ruff check src/
Related Projects
- RegenNexus UAP - Universal Adapter Protocol
- MCP Specification - Model Context Protocol
License
MIT License
常见问题
io.github.ReGenNow/regennexus-mcp 是什么?
面向 IoT 的 AI 控制硬件 MCP,支持 GPIO、robotics、sensors、cameras 与 mesh networking 等能力。
相关 Skills
网页构建器
by anthropics
面向复杂 claude.ai HTML artifact 开发,快速初始化 React + Tailwind CSS + shadcn/ui 项目并打包为单文件 HTML,适合需要状态管理、路由或多组件交互的页面。
✎ 在 claude.ai 里做复杂网页 Artifact 很省心,多组件、状态和路由都能顺手搭起来,React、Tailwind 与 shadcn/ui 组合效率高、成品也更精致。
前端设计
by anthropics
面向组件、页面、海报和 Web 应用开发,按鲜明视觉方向生成可直接落地的前端代码与高质感 UI,适合做 landing page、Dashboard 或美化现有界面,避开千篇一律的 AI 审美。
✎ 想把页面做得既能上线又有设计感,就用前端设计:组件到整站都能产出,难得的是能避开千篇一律的 AI 味。
网页应用测试
by anthropics
用 Playwright 为本地 Web 应用编写自动化测试,支持启动开发服务器、校验前端交互、排查 UI 异常、抓取截图与浏览器日志,适合调试动态页面和回归验证。
✎ 借助 Playwright 一站式验证本地 Web 应用前端功能,调 UI 时还能同步查看日志和截图,定位问题更快。
相关 MCP Server
GitHub
编辑精选by GitHub
GitHub 是 MCP 官方参考服务器,让 Claude 直接读写你的代码仓库和 Issues。
✎ 这个参考服务器解决了开发者想让 AI 安全访问 GitHub 数据的问题,适合需要自动化代码审查或 Issue 管理的团队。但注意它只是参考实现,生产环境得自己加固安全。
Context7 文档查询
编辑精选by Context7
Context7 是实时拉取最新文档和代码示例的智能助手,让你告别过时资料。
✎ 它能解决开发者查找文档时信息滞后的问题,特别适合快速上手新库或跟进更新。不过,依赖外部源可能导致偶尔的数据延迟,建议结合官方文档使用。
by tldraw
tldraw 是让 AI 助手直接在无限画布上绘图和协作的 MCP 服务器。
✎ 这解决了 AI 只能输出文本、无法视觉化协作的痛点——想象让 Claude 帮你画流程图或白板讨论。最适合需要快速原型设计或头脑风暴的开发者。不过,目前它只是个基础连接器,你得自己搭建画布应用才能发挥全部潜力。