wshobson/maverick-mcp

What it solves

MaverickMCP provides a way for AI assistants to perform professional-grade stock market analysis, portfolio tracking, and strategy backtesting locally. It bridges the gap between a general-purpose LLM and real-time financial data, allowing users to manage their investments and test trading hypotheses without needing to manually export data or use complex financial software.

How it works

It is built as a Model Context Protocol (MCP) server, which allows AI clients (like Claude Desktop, Cursor, or VS Code) to call specific tools for financial operations. It uses yfinance to fetch market data without requiring an API key for core features. The server stores user portfolios, watchlists, and trade journals in a local SQLite or PostgreSQL database, ensuring data privacy. Optional extensions add advanced capabilities like Python-based backtesting and LLM-powered financial research with web search integration.

Who it’s for

Investors, traders, and students of finance who want to integrate stock analysis and portfolio management into their AI-driven workflows. It is designed for educational and informational use rather than automated execution.

Highlights

• Comprehensive Market Data: Access quotes, historical prices, fundamentals, and technical indicators like RSI and MACD.
• Local Portfolio Management: Track positions, calculate P&L, manage risk, and maintain a trade journal locally.
• Stock Screening: Run bullish, bearish, and supply/demand screens over a custom universe of symbols.
• Advanced Backtesting: Test trading strategies using historical data, including Monte Carlo simulations and ML-based predictors.
• AI-Driven Research: Perform deep financial research on companies and sectors with source citations using integrated LLMs and web search.