For Stock Market Analysis

A production-grade Model Context Protocol (MCP) server for stock market analysis, designed to work with Claude Desktop as the AI client. Includes a local Flask web dashboard for visual analysis.

Architecture

graph TD
    subgraph Client Layer
        CD[Claude Desktop]
        FD[Flask Web Dashboard]
    end

    subgraph MCP Layer
        MS[Python MCP Server]
    end

    subgraph Tools
        FS[fetch_stock]
        GC[get_company]
        ST[statistics]
    end

    CD -- stdio --> MS
    FD -- HTTP API --> MS
    
    MS --> FS
    MS --> GC
    MS --> ST
    
    FS -- yfinance --> API[Yahoo Finance API]
    GC -- yfinance --> API

Dashboard Screenshots

Features

MCP Tools (used by Claude Desktop)

• fetch_stock — Fetch OHLCV price history for any stock symbol
• get_company — Company info: sector, market cap, P/E, 52-week range, etc.
• statistics — Full technical analysis: SMA, EMA, RSI, MACD, Bollinger Bands, Volatility, Support/Resistance

Web Dashboard

• 🌗 Premium dark theme with glassmorphism
• 📊 Interactive price charts (Chart.js)
• 📈 RSI gauge, MACD display, Bollinger Bands visualization
• 🔍 Stock search with popular picks
• ⚡ Real-time data fetching

Quick Start

1. Setup Python Environment

cd "MCP for Stock Market Analysis"
python -m venv venv
venv\Scripts\activate          # Windows
# source venv/bin/activate    # Mac/Linux
pip install -r requirements.txt

2. Run the Flask Dashboard

python flask_app/app.py

Open http://localhost:5000 in your browser.

3. Connect to Claude Desktop

Add the following to your Claude Desktop config file:

Windows: %APPDATA%\Claude\claude_desktop_config.json Mac: ~/Library/Application Support/Claude/claude_desktop_config.json

{
    "mcpServers": {
        "stock-market-analysis": {
            "command": "C:\\path\\to\\your\\project\\venv\\Scripts\\python.exe",
            "args": ["C:\\path\\to\\your\\project\\mcp_server\\server.py"]
        }
    }
}

Important: You must replace C:\path\to\your\project\ with the actual absolute path to where you saved this folder on your computer. Make sure to use double backslashes \\ in the JSON!

Restart Claude Desktop. You should see the stock market tools available.

4. Use with Claude Desktop

Ask Claude things like:

• "Fetch the stock data for AAPL for the last 3 months"
• "What are the technical indicators for Tesla?"
• "Give me a full analysis of NVDA including RSI, MACD, and Bollinger Bands"
• "Compare the volatility of AAPL vs MSFT"

Claude will automatically use the MCP tools to fetch data and provide analysis.

Project Structure

MCP for Stock Market Analysis/
├── mcp_server/                  # MCP Server (Python)
│   ├── server.py                # Main MCP server entry point
│   ├── tools/
│   │   ├── stock_fetcher.py     # Tool 1: Stock data fetcher
│   │   └── statistics_tool.py   # Tool 2: Technical indicators
│   └── database/
│       └── db.py                # SQLite caching layer
│
├── flask_app/                   # Web Dashboard
│   ├── app.py                   # Flask application
│   ├── templates/
│   │   └── index.html           # Dashboard HTML
│   └── static/
│       ├── css/style.css        # Premium dark theme
│       └── js/app.js            # Dashboard logic
│
├── requirements.txt             # Python dependencies
├── claude_desktop_config.json   # Claude Desktop config snippet
├── .env.example                 # Environment template
├── .gitignore
└── README.md

Technical Indicators

Indicator	Description	Signal
SMA (10, 20, 50)	Simple Moving Average	Price above SMA = Bullish
EMA (12, 26)	Exponential Moving Average	EMA12 > EMA26 = Bullish
RSI (14)	Relative Strength Index	<30 Oversold, >70 Overbought
MACD (12, 26, 9)	Moving Average Convergence Divergence	Histogram > 0 = Bullish
Bollinger Bands (20, 2σ)	Volatility bands	Position in band
Volatility	Standard deviation of returns	Annual risk level
Support/Resistance	Key price levels	20-day high/low

API Endpoints (Flask Dashboard)

Endpoint	Method	Description
/api/stock/<symbol>	GET	Fetch OHLCV data
/api/company/<symbol>	GET	Company information
/api/statistics/<symbol>	GET	Technical indicators
/api/health	GET	Health check

Query parameters: period (1d, 5d, 1mo, 3mo, 6mo, 1y, 5y), interval (1d, 1wk, 1mo)

Tech Stack

• MCP Server: Python, mcp[cli] SDK
• Data: yfinance, pandas, numpy
• Dashboard: Flask, Chart.js, vanilla CSS
• Database: SQLite (auto-created)
• AI Client: Claude Desktop (free tier)

License

MIT