wshobson/maverick-mcp

What it solves

MaverickMCP provides a comprehensive suite of financial analysis tools for individual traders and investors. It eliminates the need for complex setup or expensive API subscriptions by providing a local server that integrates directly with AI assistants (via the Model Context Protocol), allowing users to perform technical analysis, stock screening, and portfolio tracking through natural language.

How it works

It operates as a FastMCP server that can be connected to MCP-compatible clients like Claude Desktop, Cursor, or VS Code. The server uses yfinance for free market data and implements a tiered caching system (memory, Redis, or SQLite) to optimize performance. It offers three levels of functionality:

• Core: 37 tools for market data, technical indicators (RSI, MACD), screening, and portfolio management.
• Backtesting: An optional extension using VectorBT for strategy optimization, Monte Carlo simulations, and ML-enhanced strategies.
• Research: An optional extension using LangGraph and Exa web search for deep company and sentiment analysis (requires a BYOK LLM).

Who it’s for

Individual traders and investors who want to augment their AI assistants with real-time financial data and professional-grade analysis tools without relying on a hosted service.

Highlights

• Zero-API Core: Core market data tools require no API keys.
• Broad Client Support: Works with any MCP client via STDIO or Streamable HTTP.
• Hedge-Fund Grade Tools: Includes risk dashboards, correlation matrices, and regime-adjusted position sizing.
• Advanced Strategy Testing: Supports walk-forward analysis and ML-based signal predictors via the backtesting extra.
• Deep Research: Integrates web search and LLMs for comprehensive sector and company research.