industrial-mcp

industrial-mcp

An MCP server that lets AI models read and control industrial devices via standardized protocols like Modbus, OPC UA, and MQTT, with simulation and real hardware modes.

Category
Visit Server

README

industrial-mcp

工业设备的 USB-C 接口 — 让 AI 模型通过标准化协议读取和控制工业设备。

Python License MCP uv


解决的问题

在工业自动化领域,AI 模型(如 Claude)想要与工厂设备交互面临三大障碍:

  1. 协议碎片化 — Modbus、OPC UA、MQTT 各有各的驱动和 API
  2. 缺乏标准化 AI 接口 — 每个设备都需要定制集成代码(M×N 问题)
  3. 调试困难 — stdio 模式下 stdout 被 JSON-RPC 独占,传统 print() 调试完全失效

industrial-mcp 通过 Model Context Protocol (MCP) 解决这些问题: 把工厂里的每台设备、每个传感器、每个执行器都变成 AI 可以直接理解和操作的标准接口。


功能特性

  • 🔌 5 个 MCP Toolslist_devicesread_deviceset_speedstart_devicestop_device
  • 📡 2 个 MCP Resourcesdevice://{id}/statusplant://overview
  • 📋 2 个 MCP Prompts — 设备巡检、紧急停机
  • 🏭 3 种工业协议 — Modbus TCP、OPC UA、MQTT
  • 🎮 双模式运行 — 仿真模式(无需硬件)+ 真实设备模式
  • 🔗 物理关联模拟 — 电机转速 ↑ → 泵流量 ↑ → 传感器读数 ↑,自动传播
  • 🐛 完整调试工具链 — MCP 报文拦截器、结构化日志、健康检查端点
  • 🤖 Claude Desktop 集成 — 用自然语言控制设备:"启动传送带"、"把温度调到 80°C"
  • 🐳 Docker 支持 — 一键部署到生产环境

系统架构

graph TD
    CLAUDE[Claude Desktop / AI Model] -->|stdio JSON-RPC| SERVER[MCP Server]
    
    subgraph "industrial-mcp"
        SERVER --> DM[DeviceManager]
        DM -->|mode: simulation| SIM[SimulatorAdapter]
        DM -->|mode: real| MOD[ModbusAdapter]
        DM -->|mode: real| MQTT[MQTTAdapter]
        
        SIM --> DEV[BaseDevice subclasses]
        DEV --> MOTOR[Motor Physics]
        DEV --> PUMP[Pump Physics]
        DEV --> SENSOR[Sensor Models]
        
        MOD -->|pymodbus| HW1[Real Hardware]
        MQTT -->|paho-mqtt| HW2[Real MQTT Broker]
    end
    
    MOTOR -.->|correlated_devices| PUMP
    PUMP -.->|correlated_devices| SENSOR

快速开始

前置要求

  • Python 3.10+
  • uv 包管理器

安装

git clone https://github.com/your-org/industrial-mcp.git
cd industrial-mcp
uv sync

运行(仿真模式 — 无需硬件)

# 直接启动 MCP 服务器
uv run python -m industrial_mcp.server

# 或使用交互式演示脚本
uv run python examples/demo_simulation.py

运行(MCP Inspector 测试)

npx @modelcontextprotocol/inspector uv run python -m industrial_mcp.server

浏览器打开后 → Connect → 点击 list_devices → 看到 8 台设备。

Claude Desktop 集成

编辑配置文件:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "industrial-mcp": {
      "command": "uv",
      "args": [
        "run",
        "--directory", "/absolute/path/to/industrial-mcp",
        "python", "-m", "industrial_mcp.server"
      ]
    }
  }
}

完全退出 Claude Desktop(Cmd+Q)再重启,然后在对话中输入:

列出所有工业设备

详细配置指南 → docs/claude-desktop-setup.md


自然语言控制示例

配置完成后,直接在 Claude Desktop 中用自然语言交互:

你对 Claude 说 实际效果
"列出所有设备" 调用 list_devices(),展示 8 台设备及其状态
"启动传送带电机" 调用 start_device("motor_01_modbus"),电机开始运转
"把传送带调到 1200 RPM" 调用 set_speed("motor_01_modbus", 1200)
"检查冷却泵的温度和振动" 调用 read_device("pump_01_opcua"),返回实时数据
"做一次全面巡检" 加载 device_checkup prompt,逐台检查并生成报告
"紧急停机!" 加载 emergency_stop prompt,按优先级关闭所有设备

项目结构

industrial-mcp/
├── pyproject.toml                 # 项目配置、依赖、入口点
├── LICENSE                        # MIT 许可证
├── README.md                      # 项目说明(本文件)
│
├── config/
│   └── devices.yaml               # 设备定义(仿真 + 真实硬件配置)
│
├── src/industrial_mcp/
│   ├── server.py                  # MCP 服务器入口(Tools / Resources / Prompts)
│   ├── device_manager.py          # 设备管理器(工厂模式 + 生命周期)
│   ├── health.py                  # HTTP 健康检查端点
│   ├── mcp_logger.py              # JSON-RPC 报文拦截器(调试用)
│   │
│   ├── devices/                   # 模拟设备层(物理仿真)
│   │   ├── base.py                #    设备基类 + 全局注册表
│   │   ├── modbus_device.py       #    Modbus 模拟设备
│   │   ├── opcua_device.py        #    OPC UA 模拟设备
│   │   └── mqtt_device.py         #    MQTT 模拟设备
│   │
│   ├── protocols/                 # 协议适配器层(统一接口)
│   │   ├── base.py                #    ProtocolAdapter 抽象基类 + 重试逻辑
│   │   ├── simulator.py           #    仿真适配器(包装 BaseDevice)
│   │   ├── modbus.py              #    Modbus TCP 适配器(pymodbus)
│   │   └── mqtt.py               #    MQTT 适配器(paho-mqtt)
│   │
│   ├── models/                    # Pydantic 数据模型
│   │   ├── device.py              #    设备配置、寄存器映射
│   │   └── telemetry.py           #    遥测数据(电机/泵/传感器)
│   │
│   └── utils/                     # 工具
│       ├── logger.py              #    结构化日志(stderr + 文件轮转)
│       └── logging.py             #    日志配置
│
├── docs/                          # 文档
│   ├── architecture.md            #    架构设计(含 Mermaid 图)
│   ├── api.md                     #    MCP API 参考文档
│   ├── claude-desktop-setup.md    #    Claude Desktop 集成指南
│   ├── debugging-guide.md         #    调试指南
│   └── deployment.md              #    部署指南(Docker / systemd)
│
└── examples/                      # 示例
    ├── demo_simulation.py         #    交互式设备仿真演示
    ├── mcp_client_demo.py         #    MCP 客户端编程示例
    ├── claude_desktop_config.json #    Claude Desktop 配置模板
    └── claude_desktop_config_debug.json  # 调试配置模板

技术栈

类别 技术 用途
MCP 协议 mcp SDK v2 (Python) MCP 服务器实现
数据模型 Pydantic v2 配置和遥测数据结构
工业协议 pymodbus, opcua-asyncio, paho-mqtt Modbus/OPC UA/MQTT 通信
配置 YAML (PyYAML) 声明式设备定义
日志 structlog 结构化日志(stderr + 文件)
包管理 uv (Astral) 依赖管理和虚拟环境
代码质量 ruff, mypy Lint + 类型检查
测试 pytest, pytest-asyncio 异步测试框架
部署 Docker, systemd 容器化和系统服务

贡献指南

我们欢迎 Issue、PR 和任何形式的贡献!

开发环境搭建

git clone https://github.com/your-org/industrial-mcp.git
cd industrial-mcp
uv sync --dev

代码规范

uv run ruff check src/     # Lint
uv run ruff format src/    # 格式化
uv run mypy src/           # 类型检查

提交 PR 流程

  1. Fork 本仓库
  2. 创建功能分支 (git checkout -b feat/amazing-feature)
  3. 提交更改 (git commit -m 'Add amazing feature')
  4. 推送到分支 (git push origin feat/amazing-feature)
  5. 创建 Pull Request

添加新协议适配器

  1. src/industrial_mcp/protocols/ 中创建新文件
  2. 继承 ProtocolAdapter 抽象基类
  3. 实现 connect, disconnect, read_register, write_register, read_all
  4. DeviceManager._create_adapter() 中注册
  5. config/devices.yaml 中添加配置示例

许可证

MIT License — 随意使用、修改和分发。


致谢


🤖 industrial-mcp — 让 AI 进入工厂车间,从这一行命令开始:

npx @modelcontextprotocol/inspector uv run python -m industrial_mcp.server

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured