Skip to main content
Glama
jiangyj545

A-Stock MCP Server

by jiangyj545

A-Stock MCP Server 🇨🇳📈

A Model Context Protocol (MCP) server providing Chinese A-Share stock market data — real-time quotes, historical K-lines, market scanning, and intelligent stock screening.

Zero API key required. Powered by BaoStock (free) + Sina Finance API.

Features

🛠 MCP Tools

Tool

Description

Example

get_quote

Real-time quote for one or more stocks

"茅台实时报价"

get_kline

Historical K-line data (daily/weekly/monthly/5min/15min/30min/60min)

"宁德时代最近60日K线"

get_stock_list

List all A-share stocks with filtering

"列出所有沪深300成分股"

scan_market

Market scan: top gainers, losers, volume leaders, active stocks

"今日涨幅前20"

screen_stocks

Multi-factor stock screening (price, volume, MA, sector)

"放量突破MA20的股票"

get_stock_info

Company fundamentals: industry, listing date, description

"比亚迪公司信息"

get_dividend

Dividend history for a stock

"长江电力分红记录"

Key Highlights

  • 6937+ A-share stocks covered (SH + SZ)

  • Zero cost — no API keys, no subscriptions

  • Real-time quotes via Sina API

  • Historical data up to 20+ years via BaoStock

  • Smart screening — combine price, volume, MA, and sector filters

  • Batch queries — ask for multiple stocks at once

Related MCP server: A-Share MCP Server

Installation

pip install astock-mcp-server

Option 2: uvx (no install)

uvx astock-mcp-server

Option 3: From source

git clone https://github.com/yunjie/astock-mcp-server.git
cd astock-mcp-server
pip install -e .

Configuration

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "astock": {
      "command": "uvx",
      "args": ["astock-mcp-server"]
    }
  }
}

Cursor / VS Code

Add to your MCP settings:

{
  "mcpServers": {
    "astock": {
      "command": "python",
      "args": ["-m", "astock_mcp.server"]
    }
  }
}

Hermes Agent

Add to your config.yaml:

mcp:
  servers:
    astock:
      command: "uvx"
      args: ["astock-mcp-server"]

Usage Examples

Once connected, your AI assistant can:

"帮我查一下贵州茅台和宁德时代的实时报价"
→ Uses get_quote with codes ["600519.SH", "300750.SZ"]

"找出今天涨幅超过5%且成交额超过10亿的股票"
→ Uses scan_market with filters

"比亚迪最近一年的周K线数据"
→ Uses get_kline with frequency="w" and periods=52

"筛选价格在20-50元之间、放量突破20日均线、属于新能源板块的股票"
→ Uses screen_stocks with multi-factor filters

Data Sources

Source

Data

Cost

Coverage

BaoStock

Historical K-lines, stock list, fundamentals, dividends

Free

1990-present

Sina Finance

Real-time quotes, market rankings

Free

All A-shares

Stock Code Format

Use standard A-share codes with exchange suffix:

  • Shanghai (上交所): 600519.SH, 601318.SH

  • Shenzhen (深交所): 000858.SZ, 300750.SZ

  • Beijing (北交所): 872925.BJ

Limitations

  • Real-time quotes have ~3 second delay (not for HFT)

  • BaoStock data updates after market close (no intraday historical yet)

  • Rate limiting: BaoStock ~100 requests/min, Sina ~60 requests/min

License

MIT — use freely, no warranty.

Contributing

Issues and PRs welcome! This is an open-source project.

Related MCP Connectors

Related MCP Servers

  • A
    license
    B
    quality
    C
    maintenance
    Provides comprehensive access to Chinese stock market data including real-time quotes, historical prices, financial statements, technical indicators, news, and macroeconomic data for A/B/H shares through multiple data sources.
    30
    46
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    Provides comprehensive A-share (Chinese stock market) data including stock information, historical prices, financial reports, macroeconomic indicators, technical analysis, and valuation metrics through the free Baostock data source.
    27
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    Enables users to query Chinese A-share stock data, including stock basics, historical prices, financial reports, and macroeconomic indicators.
    1
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    Provides free, open-source access to China A-share financial data for quantitative analysis, including stock metrics, news, and backtesting.
    28
    MIT