This project provides a professional financial data server (MCP) designed to offer comprehensive, real-time financial data access to AI assistants like Claude. It integrates Tushare and Binance APIs, enabling multi-dimensional analysis across stocks, funds, bonds, cryptocurrencies, macroeconomics, and news. The primary benefit is empowering AI models with robust financial data capabilities for sophisticated analysis and querying.

How It Works

FinanceMCP acts as a middleware, translating natural language queries into structured API calls to Tushare and Binance. It employs an intelligent technical indicator system that automatically prefetches necessary historical data and enforces strict parameterization (e.g., macd(12,26,9)) for precise calculations. The system features a modular architecture for parameter parsing, data fetching, and indicator computation, supporting core indicators like MACD, RSI, KDJ, BOLL, and MA. This approach ensures accuracy and efficiency in delivering complex financial metrics.

Quick Start & Requirements

Users can opt for a free public cloud service for immediate testing (http://106.14.205.176:3090/ for web, or configure Claude Desktop with http://47.79.147.241:3100/mcp). For local deployment, Node.js (v18+) and a Tushare API token are required. Installation can be done via npm (npm install -g finance-mcp) or manual cloning from GitHub. The recommended setup for local use is the stdio mode (npx -y finance-mcp), offering low latency and resource efficiency. Configuration details for Claude Desktop are provided for both stdio and HTTP modes.

Highlighted Details

• Market Coverage: Supports 10 markets including A-shares, US stocks, HK stocks, Forex, Futures, Funds, Bonds, Options, and cryptocurrencies via Binance.
• Data Breadth: Offers real-time news, 11 macro-economic indicators, detailed company financials (A-share, HK, US), fund data, convertible bonds, block trades, money flow analysis, margin trading data, and dragon-tiger lists.
• Intelligent Indicators: Features MACD, RSI, KDJ, BOLL, MA with automatic data prefetching and strict parameter enforcement.
• Performance: Fund data query performance is claimed to be improved by 85% (from 5.2s to 0.8s).
• AI Native: Designed for seamless integration with AI assistants, supporting natural language queries.

Maintenance & Community

The project is maintained by 陈星宇 (Chen Xingyu), with contact via email (guangxiangdebizi@gmail.com) and GitHub. No specific community channels (like Discord/Slack) or formal roadmap links are provided in the README.

Licensing & Compatibility

The project is released under the MIT License, which is permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

The public cloud services are noted as "personal small servers" and users are advised against abuse. Full, unlimited access to data, especially for production environments, requires obtaining and configuring a personal Tushare API token. The system's functionality is dependent on the availability and data provided by the Tushare and Binance APIs.