MaverickMCP

# MaverickMCP Database Setup This guide explains how to set up and seed the SQLite database for MaverickMCP with sample stock data. ## Quick Start ### 1. Run Complete Setup (Recommended) ```bash # Set your database URL (optional - defaults to SQLite) export DATABASE_URL=sqlite:///maverick_mcp.db # Run the complete setup script ./scripts/setup_database.sh ``` This will: - ✅ Create SQLite database with all tables - ✅ Seed with 40 sample stocks (AAPL, MSFT, GOOGL, etc.) - ✅ Populate with 1,370+ price records - ✅ Generate sample screening results (Maverick, Bear, Supply/Demand) - ✅ Add technical indicators cache ### 2. Manual Step-by-Step Setup ```bash # Step 1: Create database tables python scripts/migrate_db.py # Step 2: Seed with sample data (no API key required) python scripts/seed_db.py # Step 3: Test the setup python scripts/test_seeded_data.py ``` ## Database Configuration ### Default Configuration (SQLite) - **Database**: `sqlite:///maverick_mcp.db` - **Location**: Project root directory - **No setup required**: Works out of the box ### PostgreSQL (Optional) ```bash # Set environment variable export DATABASE_URL=postgresql://localhost/maverick_mcp # Create PostgreSQL database createdb maverick_mcp # Run migration python scripts/migrate_db.py ``` ## Sample Data Overview ### Stocks Included (40 total) - **Large Cap**: AAPL, MSFT, GOOGL, AMZN, TSLA, NVDA, META, BRK-B, JNJ, V - **Growth**: AMD, CRM, SHOP, ROKU, ZM, DOCU, SNOW, PLTR, RBLX, U - **Value**: KO, PFE, XOM, CVX, JPM, BAC, WMT, PG, T, VZ - **Small Cap**: UPST, SOFI, OPEN, WISH, CLOV, SPCE, LCID, RIVN, BYND, PTON ### Generated Data - **1,370+ Price Records**: 200 days of historical data for 10 stocks - **24 Maverick Stocks**: Bullish momentum recommendations - **16 Bear Stocks**: Bearish setups with technical indicators - **16 Supply/Demand Breakouts**: Accumulation breakout candidates - **600 Technical Indicators**: RSI, SMA cache for analysis ## Testing MCP Tools After seeding, test that the screening tools work: ```bash python scripts/test_seeded_data.py ``` Expected output: ``` ✅ Found 5 Maverick recommendations 1. PTON - Score: 100 2. BYND - Score: 100 3. RIVN - Score: 100 ✅ Found 5 Bear recommendations 1. MSFT - Score: 37 2. JNJ - Score: 32 3. TSLA - Score: 32 ✅ Total screening results across all categories: 56 ``` ## Using with Claude Desktop After database setup, start the MCP server: ```bash # Start the server make dev # Or manually uvicorn maverick_mcp.api.server:app --host 0.0.0.0 --port 8003 ``` Then connect with Claude Desktop using `mcp-remote`: ```json { "mcpServers": { "maverick-mcp": { "command": "npx", "args": ["-y", "mcp-remote", "http://localhost:8003/mcp"] } } } ``` Test with prompts like: - "Show me the top maverick stock recommendations" - "Get technical analysis for AAPL" - "Find bearish stocks with high RSI" ## Database Schema ### Core Tables - **mcp_stocks**: Stock symbols and company information - **mcp_price_cache**: Historical OHLCV price data - **mcp_technical_cache**: Calculated technical indicators ### Screening Tables - **mcp_maverick_stocks**: Bullish momentum screening results - **mcp_maverick_bear_stocks**: Bearish setup screening results - **mcp_supply_demand_breakouts**: Breakout pattern screening results ## Troubleshooting ### Database Connection Issues ```bash # Check database exists ls -la maverick_mcp.db # Test SQLite connection sqlite3 maverick_mcp.db "SELECT COUNT(*) FROM mcp_stocks;" ``` ### No Screening Results ```bash # Verify data was seeded sqlite3 maverick_mcp.db " SELECT (SELECT COUNT(*) FROM mcp_stocks) as stocks, (SELECT COUNT(*) FROM mcp_price_cache) as prices, (SELECT COUNT(*) FROM mcp_maverick_stocks) as maverick; " ``` ### MCP Server Connection ```bash # Check server is running curl http://localhost:8003/health # Check MCP endpoint curl http://localhost:8003/mcp/capabilities ``` ## Advanced Configuration ### Environment Variables ```bash # Database DATABASE_URL=sqlite:///maverick_mcp.db # Optional: Enable debug logging LOG_LEVEL=debug # Optional: Redis caching REDIS_HOST=localhost REDIS_PORT=6379 ``` ### Custom Stock Lists Edit `scripts/seed_db.py` and modify `SAMPLE_STOCKS` to include your preferred stock symbols. ### Production Setup - Use PostgreSQL for better performance - Enable Redis caching - Set up proper logging - Configure rate limiting --- ✅ **Database ready!** Your MaverickMCP instance now has a complete SQLite database with sample stock data and screening results.