Tushare MCP Server
Provides access to Tushare financial data through the Model Context Protocol, enabling users to query and retrieve Chinese stock market and financial information.
README
Tushare MCP Server
This is a Model Context Protocol (MCP) server that provides access to Tushare financial data.
Prerequisites
- Python 3.10 or higher
- A Tushare Token (Get it from Tushare.pro)
Installation
- Install the dependencies:
pip install -r requirements.txt
Configuration
- Create a
.envfile in the project root directory:TUSHARE_TOKEN=your_tushare_token_here
Project Layout
server/: MCP entrypoint (server.py) plus logs.src/: application code split intostrategies/(wheel backtest, etc.) andutils/(shared math + performance helpers).temp_data/: generated artifacts such aswheel_report.html,wheel_dashboard.html, and portfolio CSV/JSON outputs.tests/: standalone scripts for strategy prototyping (wheel_strategy.py,portfolio_rebalance.py, etc.).
Using with GitHub Copilot in VS Code
To use this MCP server with GitHub Copilot in VS Code, you need to configure the mcp.json file.
-
Open Configuration:
- Open the Command Palette (
Ctrl+Shift+PorF1). - Search for and select
MCP: Configure MCP Servers. - This will open the
mcp.jsonfile (typically located in%APPDATA%\Code\User\mcp.jsonon Windows).
- Open the Command Palette (
-
Add via Command Palette (quick way):
- Press
Ctrl+Shift+Pagain and pickMCP: Add MCP Server. - When the Enter Command prompt appears, paste:
```
C:\Users\lochen\AppData\Local\Microsoft\WindowsApps\python3.13.exe c:\Users\lochen\tushare-mcp\server\server.py
```
- Accept the suggested server name (for example
tushare) and save.
-
Add Server Configuration manually: Add the
tushare-serverconfiguration to the JSON file. Make sure to use absolute paths for both the Python executable and the script.{ "mcpServers": { "tushare-server": { "command": "C:\\path\\to\\your\\python.exe", "args": [ "C:\\path\\to\\tushare_mcp_server\\server\\server.py" ] } } }- Replace
C:\\path\\to\\your\\python.exewith your actual Python interpreter path (e.g.,C:\\Users\\username\\AppData\\Local\\Programs\\Python\\Python311\\python.exe). - Replace
C:\\path\\to\\tushare_mcp_server\\server\\server.pywith the absolute path to this project's server entry point.
- Replace
-
Restart VS Code: After saving
mcp.json, restart VS Code for the changes to take effect.
Usage
Testing with MCP Inspector
You can test the server using the MCP Inspector:
npx @modelcontextprotocol/inspector python server/server.py
Price Volatility Tool
After the server is running (Inspector, Copilot, etc.), invoke the get_price_volatility tool to compute recent volatility for a stock. It supports frequency values daily, monthly, or yearly, letting you measure 波动率 on日线/月线/年线 data.
{
"tool": "get_price_volatility",
"args": {
"identifier": "000001.SZ",
"window": 30,
"frequency": "daily",
"annualize": true
}
}
Key fields returned:
frequency: 数据频率(daily/monthly/yearly)。window_periods: 实际使用的周期数(交易日/月份/年份)。period_volatility: 该频率下的标准差。annualized_volatility: 根据频率自动换算后的年化波动率(252/12/1)。mean_period_return: 平均单周期收益。
Option Reference & Daily Data
基于 Tushare 文档 #157 新增了两个期权工具:
get_option_basic(exchange, fields): 返回上/深交所上市期权的合约元数据(行权价、类型、到期日等)。get_option_daily(ts_code, trade_date, start_date, end_date, exchange): 返回期权日线行情。
示例(通过 MCP 调用 get_option_daily 查询科创50期权在单日的 K 线):
{
"tool": "get_option_daily",
"args": {
"exchange": "SSE",
"trade_date": "20251203"
}
}
返回值为 JSON 数组,字段与 Tushare opt_daily 接口一致,可配合 get_price_volatility 等工具进一步分析期权策略。
ETF/Fund Daily Quotes
使用新的 get_fund_daily 工具(Tushare 文档 #127 对应接口)可直接拉取 ETF/场内基金的日线行情:
{
"tool": "get_fund_daily",
"args": {
"ts_code": "159915.SZ",
"start_date": "20250101",
"end_date": "20251203",
"fields": "ts_code,trade_date,open,high,low,close,vol"
}
}
当 fields 为空时会返回默认全部列。可搭配 get_option_daily、get_price_volatility 等工具构建 ETF+期权策略分析。
Wheel Strategy Backtest
tests/wheel_strategy.py 利用 159915.SZ(创业板ETF)及其期权,ETF 行情通过 Tushare fund_daily 接口(文档 #127)获取,模拟“车轮饼”策略:
- 每个自然月首个交易日:
- 若空仓,卖出 5%~10% OTM 的认沽合约(
call_put='P')。 - 若持仓,卖出 5%~10% OTM 的认购合约(
call_put='C')。
- 到期日根据标的收盘价判断是否被指派,按轮动逻辑更新持仓。
- 统计权利金、被指派次数、最大保证金占用,并估算收益率。
- 报告会显示每笔交易的隐含波动率:若 Tushare 返回该字段则直接使用,否则基于 Black-Scholes(假定 2% 无风险利率)用成交价反推出一个参考值。
运行:
python tests/wheel_strategy.py
输出包含总收益、占用保证金、近10期交易记录等。所有生成的 wheel_report.html、wheel_dashboard.html 等文件都会写入 temp_data/,便于统一清理。若需调整标的、时间区间或虚值区间,可编辑脚本顶部的常量(UNDERLYING、START_DATE、OTM_RANGE 等)。
Wheel Strategy MCP Tool
无需运行脚本,也可直接通过 MCP 调用 backtest_wheel_strategy:
{
"tool": "backtest_wheel_strategy",
"args": {
"underlying": "159915.SZ",
"start_date": "20230101",
"end_date": "20251203",
"otm_min": 0.05,
"otm_max": 0.10,
"initial_capital": 30000
}
}
返回 JSON 中包含:
ending_value、return_on_capital、annualized_return等指标。recent_trades:近 12 期的期权选择、权利金、行权价及是否被指派。
可通过参数更换标的(只要该 ETF 有挂牌期权)、调整虚值区间或回测时间窗。确保 Tushare Token 对 fund_daily、opt_basic、opt_daily 接口有权限。
Multi-ETF Portfolio Backtest
tests/portfolio_rebalance.py 按照截图中的 10 只 ETF 及固定权重构建组合,并在每个自然月首个交易日动态再平衡:
python tests/portfolio_rebalance.py
脚本会:
- 自动获取所有 ETF 的可用历史区间,并截取重叠部分;
- 计算每日组合净值并输出
temp_data/portfolio_equity_curve.csv,再平衡明细写入temp_data/portfolio_rebalances.csv; - 生成
temp_data/portfolio_vs_benchmarks.csv,其中包含组合与沪深300/中证500/创业板指的归一化指数曲线; - 在
temp_data/portfolio_summary.json中汇总收益率、年化波动率、最大回撤、Sharpe Ratio 及各基准指数的对比指标。
Using with Claude Desktop
Add the following configuration to your claude_desktop_config.json:
{
"mcpServers": {
"tushare": {
"command": "python",
"args": ["C:\\Users\\lochen\\tushare_mcp_server\\server\\server.py"],
"env": {
"TUSHARE_TOKEN": "your_token_here"
}
}
}
}
Make sure to replace your_token_here with your actual Tushare token.
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.