Tushare MCP Server

A Model Context Protocol (MCP) server that provides AI assistants with access to Tushare financial data API. This enables AI models to query Chinese stock market data, financial statements, macroeconomic indicators, and more.

Features

This MCP server exposes 18 financial data tools across multiple categories:

Stock Data

• stock_basic - Stock list with basic information (code, name, industry, listing date)
• stock_company - Company profile and basic information
• daily_quote - Daily price data (OHLCV)
• daily_basic - Daily valuation metrics (PE, PB, PS, market cap)
• trade_cal - Trading calendar

Financial Statements

• income - Income statement data
• balance_sheet - Balance sheet data
• cashflow - Cash flow statement data
• fina_indicator - Financial indicators (ROE, gross margin, net margin)

Market Data

• daily_limit - Daily limit-up/limit-down statistics
• moneyflow_hsgt - Stock Connect (Northbound) capital flow
• top10_holders - Top 10 shareholders

Concept & Index

• concept - Concept stock categories
• concept_detail - Concept stock constituents
• index_basic - Index basic information
• index_daily - Index daily quotes

Futures

• future_basic - Futures contract basic info
• future_daily - Futures daily market data

Installation

From Source

git clone https://github.com/lijinly/tushare-mcp-server.git
cd tushare-mcp-server
pip install -e .

Requirements

• Python >= 3.10
• Tushare API Token (Get one free)

Configuration

1. Get Tushare API Token

Register at Tushare Pro and get your API token from your profile page.

2. Configure Environment Variable

# Linux/macOS
export TUSHARE_TOKEN="your_token_here"

# Windows PowerShell
$env:TUSHARE_TOKEN="your_token_here"

# Windows CMD
set TUSHARE_TOKEN=your_token_here

3. Configure MCP Client

Add to your MCP client configuration (e.g., Claude Desktop, Cline):

{
  "mcpServers": {
    "tushare": {
      "command": "python",
      "args": ["-m", "tushare_mcp_server"],
      "env": {
        "TUSHARE_TOKEN": "your_token_here"
      }
    }
  }
}

Claude Desktop Config Location:

• Windows: %APPDATA%\Claude\claude_desktop_config.json
• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

Usage

Manual Start

python -m tushare_mcp_server

With MCP Client

Once configured, the AI assistant can directly call Tushare data tools:

Example queries:

• "Get the daily quote for 000001.SZ"
• "Show me the PE ratio of Kweichow Moutai"
• "What's the stock connect capital flow today?"
• "List all concept stocks for artificial intelligence"

Data Format

All tools return JSON formatted data:

{
  "data": [...],
  "count": 100,
  "columns": ["ts_code", "trade_date", "close", "pe", "pb"]
}

Project Structure

tushare-mcp-server/
├── src/tushare_mcp_server/
│   ├── __init__.py
│   ├── __main__.py
│   └── server.py          # Main MCP server implementation
├── pyproject.toml         # Project configuration
├── README.md
└── .gitignore

Dependencies

Package	Version	Purpose
mcp	>=1.0.0	MCP protocol implementation
tushare	>=1.4.0	Tushare API client
pandas	>=2.0.0	Data manipulation

Development

Setup Development Environment

# Create virtual environment
conda create -n tushare-mcp python=3.10 -y
conda activate tushare-mcp

# Install in editable mode
pip install -e .

Testing

python -m tushare_mcp_server

Troubleshooting

"TUSHARE_TOKEN environment variable is not set"

Make sure you've set the TUSHARE_TOKEN environment variable before starting the server.

"Error connecting to Tushare API"

• Check your internet connection
• Verify your Tushare token is valid
• Ensure your Tushare account has sufficient API points

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Tushare - Chinese financial data platform
• Model Context Protocol - Open protocol for AI tool integration

Related Links

• Tushare API Documentation
• MCP Specification