MCP Server: Technical Analysis - PDF & Web Report

A comprehensive technical analysis platform that combines FastAPI endpoints, Model Context Protocol (MCP) integration, and a Flask frontend:

• Three endpoint types:
  • / - Flask-based web interface
  • /api/* - FastAPI endpoints for direct programmatic access
  • /mcp - Model Context Protocol endpoint for AI/LLM integration
• Enhanced with MCP capabilities using the fastapi-mcp package
• Detailed API documentation with Pydantic models
• Both synchronous and asynchronous processing
• Generates professional PDF and HTML reports

Endpoints

Access Methods

• Web Interface: visit / for the Flask frontend
• Direct API: make HTTP requests to /api/technical-analysis
• MCP/LLM: connect to /mcp using any MCP-compatible client

Refer to the GitHub repository README for detailed information about request formats, response structures, and example usage.

Data Collection

• Historical price data fetched from Yahoo Finance API
• Supports both daily and weekly timeframes
• Uses the custom Yahoo Finance FastAPI MCP server (yfin-h.tigzig.com)

Technical Analysis

• Computes EMAs, MACD, RSI, Bollinger Bands
• Custom chart generation for both timeframes
• Pattern recognition algorithms

AI Integration

• Data and charts processed by Google's Gemini Vision API
• Comprehensive market interpretation in plain English
• MCP exposed to any AI/LLM client

1. Clone the repository
2. Install dependencies: pip install -r requirements.txt
3. Configure environment variables:

   GEMINI_API_KEY=your_gemini_api_key
   GEMINI_MODEL_NAME=gemini-2.5-flash

4. Run the server: uvicorn main:app --host 0.0.0.0 --port 8000 --timeout-keep-alive 900

Key Dependencies

• MCP Server: Yahoo Finance (companion service)
• FastAPI Server: ReportLab (for PDF generation)