MCP Zhitou Server

🚀 MCP Protocol and Zhitu Data Service API Integration Guide

This project offers an integration solution based on the MCP protocol and Zhitu Data Service API. It aims to seamlessly connect AI assistants with the Shanghai and Shenzhen stock data services provided by Zhitu Data Service through the MCP (Multi - Modal Protocol). With this solution, AI can directly access real - time stock data and market information, providing strong support for investment analysis and decision - making.

🚀 Quick Start

1. Initialize the MCP Service

In the main file of the project, initialize an MCP agent instance:

from mcp import Agent

# Create an MCP agent instance
agent = Agent(
    name="TradingAssistant",
    description="A comprehensive trading assistant integrating with MCP protocol.",
    tools=TOOLS.values()
)

2. Define Available Tools

Define the available tools in the mcp_z兔_server.py file:

# Tool list
TOOLS = {
    "get_stock_list": get_stock_list,
    "get_new_stock_calendar": get_new_stock_calendar,
    "get_company_profile": get_company_profile,
    "get_capital_daily_trend": get_capital_daily_trend,
    "get_all_announcements": get_all_announcements
}

3. Start the Service

Start the MCP agent service:

if __name__ == "__main__":
    agent.run()

✨ Features

This integration solution enables AI assistants to directly access real - time stock data and market information from Zhitu Data Service through the MCP protocol, providing strong support for investment analysis and decision - making.

📦 Installation

1. Environment Requirements

• Python Version: Python 3.6 or higher.
• MCP Protocol Library: You need to install the mcp - client library to communicate with the AI agent.

2. Installation Steps

# Install the MCP client library
pip install mcp - client

# Clone the project repository
git clone https://github.com/your - username/mcp - trading.git
cd mcp - trading

💻 Usage Examples

Basic Usage

The following shows how to initialize the MCP service, define available tools, and start the service:

from mcp import Agent

# Create an MCP agent instance
agent = Agent(
    name="TradingAssistant",
    description="A comprehensive trading assistant integrating with MCP protocol.",
    tools=TOOLS.values()
)

# Tool list
TOOLS = {
    "get_stock_list": get_stock_list,
    "get_new_stock_calendar": get_new_stock_calendar,
    "get_company_profile": get_company_profile,
    "get_capital_daily_trend": get_capital_daily_trend,
    "get_all_announcements": get_all_announcements
}

if __name__ == "__main__":
    agent.run()

Advanced Usage

If you want to add new tools, you can follow the steps in the Extension Guide. For example:

# New tool example
def new_tool_function(params):
    # Implementation logic
    pass

TOOLS["new_tool"] = new_tool_function

📚 Documentation

Tools Available

The currently implemented tools are as follows:

Extension Guide

1. Consult the Zhitu API Documentation

Visit Zhitu Data Service API Documentation to find the API endpoints for the required functions.

2. Implement New Tool Functions

In mcp_z兔_server.py, develop new Python functions to call the target API following the pattern of the existing tool functions. Ensure that you handle parameters and return values properly and add clear comments.

3. Register New Tools

Register the new tool in the TOOLS dictionary:

# New tool example
def new_tool_function(params):
    # Implementation logic
    pass

TOOLS["new_tool"] = new_tool_function

Configuration Guide

1. MCP Client Configuration

Create a configuration file config.json:

{
    "mcp": {
        "host": "localhost",
        "port": 5000,
        "timeout": 30
    },
    "api_key": "your_api_key_here"
}

Error Handling

1. Common Errors

• API call failed: Check the network connection and the validity of the API key.
• Tool not registered: Make sure all tools are correctly added to the TOOLS dictionary.

🔧 Technical Details

This project uses the MCP protocol to connect AI assistants with Zhitu Data Service. The mcp - client library is used to communicate with the AI agent, and the service is started through the Agent class. The tools are defined in the TOOLS dictionary, and the configuration information is stored in the config.json file.

📄 License

This project is licensed under the MIT License. Please refer to the LICENSE file for detailed information.

🤝 Contributing

Contributions are welcome! Please refer to CONTRIBUTING.md for ways to participate.

By following the above steps, you can easily integrate the MCP protocol with the Zhitu Data Service API, adding powerful financial data processing capabilities to your AI project.