Dingtalkmcpserver

🚀 DingTalk MCP - Invoke Locally Privately Deployed MCP Services via DingTalk Client AI Assistant

You can start and run the project by importing the repository into a Python IDE project without remote deployment.

✨ Features

• Enable direct communication between the DingTalk AI assistant's direct mode and local services.
• Support the MCP protocol with multi - model configuration.
• Support the locally run MCP Service.

📁 Project Directory Structure

[
  '|-- Project',
  '    |-- README.md',
  '    |-- main.py',
  '    |-- app',
  '        |-- api',
  '        |   |-- __init__.py',
  '        |   |-- dingtalk_client.py',
  '        |-- config',
  '        |   |-- __init__.py',
  '        |   |-- settings.py',
  '        |-- core',
  '        |   |-- __init__.py',
  '        |   |-- llm_service.py',
  '        |   |-- mcp_server.py',
  '        |   |-- message_service.py',
  '        |-- resource',
  '        |   |-- DingTalkAgentConfig',
  '        |   |-- conversation.jpg',
  '        |-- utils',
  '            |-- __init__.py',
  '            |-- logger.py',
  ''
]

🔧 Custom Function Development

In the app/core/mcp_server.py file, you can use the @register_tool decorator to add your own custom functions.

📦 Installation

Environment Requirements

• Python > 3.10
• OpenSSL > 1.1

Using the Local Environment

1. Install dependencies:

   You can directly use the IDE to import them.
   

👋 Contribution

Welcome to contribute! Feel free to communicate. Contact email: miechalzhao@gmail.com