MaverickMCP is a personal-use FastMCP 2.0 server that exposes a comprehensive stock analysis toolkit directly to MCP clients such as Claude Desktop. It provides 29+ financial tools for technical analysis, screening, backtesting, and portfolio optimization, built on Python with uv for fast dependency management. The server is pre-seeded with 520 S&P 500 stocks and offers multiple transports (HTTP, SSE, and STDIO) so you can connect via standard MCP clients or integrate into Claude Desktop for seamless tool registration and execution. By connecting through SSE or STDIO, you gain reliable tool availability, session persistence, and efficient data caching via Redis with graceful fallback to memory when Redis is not present. You can leverage backtesting, vectorized indicators (e.g., SMA, EMA, RSI, MACD), stock screening strategies, and portfolio analysis to perform end-to-end research within your MCP-enabled environment.

Prerequisites:

• Python 3.12+ installed
• uv (recommended) or pip installation of the MCP tooling
• Optional: Redis for caching, PostgreSQL/SQLite for persistence, TA-Lib for technical indicators

Option A: Install with uv (Recommended - Fastest)

1. Clone the repository git clone https://github.com/wshobson/maverick-mcp.git
2. Enter the project cd maverick-mcp
3. Install dependencies and set up using uv uv sync
4. Copy environment template and configure API keys cp .env.example .env

   Add your Tiingo API key in the .env file

5. Start the server (per the README, this is typically via the project Makefile, e.g., make dev)

Option B: Traditional pip install (if you prefer a standard Python workflow)

1. Clone the repository git clone https://github.com/wshobson/maverick-mcp.git
2. Create a virtual environment and activate it python -m venv .venv

   On macOS/Linux

   source .venv/bin/activate

   On Windows

   .venv\Scripts\activate
3. Install in editable mode pip install -e .
4. Copy environment template cp .env.example .env

   Add your Tiingo API key in the .env file

5. Start the server (as documented in the Makefile): make dev

Connecting to Claude Desktop (example configuration) is described in the README and can be adapted to your setup.

Tips and notes:

• The server runs with multiple transport options; choose SSE for stable tool registration or STDIO for development/testing.
• Ensure your Tiingo API key is set in the environment (.env) to enable data access; without it you may run with limited data.
• Redis is optional but recommended for caching; if Redis is unavailable, the system will gracefully fall back to in-memory caching.
• The default database backend uses SQLAlchemy with PostgreSQL/SQLite, defaulting to SQLite if PostgreSQL is not configured.
• If you plan to connect Claude Desktop, use the provided mcp-remote JSON snippet with the trailing slash on the SSE URL for reliability.
• After changing Claude Desktop configuration, restart the application to ensure tool registrations are reloaded correctly.
• When using TA-Lib, follow the platform-specific installation instructions provided in the README to ensure the library compiles and links correctly.