MaverickMCP

"""MCP Introspection Tools for Better Discovery and Understanding.""" from typing import Any from fastmcp import FastMCP def register_introspection_tools(mcp: FastMCP) -> None: """Register introspection tools for better discovery.""" @mcp.tool(name="discover_capabilities") async def discover_capabilities() -> dict[str, Any]: """ Discover all available capabilities of the MaverickMCP server. This tool provides comprehensive information about: - Available strategies (traditional and ML) - Tool categories and their functions - Parameter requirements for each strategy - Example usage patterns Use this as your first tool to understand what's available. """ return { "server_info": { "name": "MaverickMCP", "version": "1.0.0", "description": "Advanced stock analysis and backtesting MCP server", }, "capabilities": { "backtesting": { "description": "Run and optimize trading strategies", "strategies_available": 15, "ml_strategies": ["online_learning", "regime_aware", "ensemble"], "traditional_strategies": [ "sma_cross", "rsi", "macd", "bollinger", "momentum", "ema_cross", "mean_reversion", "breakout", "volume_momentum", ], "features": [ "parameter_optimization", "strategy_comparison", "walk_forward_analysis", ], }, "technical_analysis": { "description": "Calculate technical indicators and patterns", "indicators": [ "SMA", "EMA", "RSI", "MACD", "Bollinger Bands", "Support/Resistance", ], "chart_analysis": True, "pattern_recognition": True, }, "screening": { "description": "Pre-calculated S&P 500 screening results", "strategies": [ "maverick_bullish", "maverick_bearish", "supply_demand_breakouts", ], "database": "520 S&P 500 stocks pre-seeded", }, "research": { "description": "AI-powered research with parallel execution", "features": [ "comprehensive_research", "company_analysis", "sentiment_analysis", ], "performance": "7-256x speedup with parallel agents", "ai_models": "400+ models via OpenRouter", }, }, "quick_start": { "first_command": "Run: discover_capabilities() to see this info", "simple_backtest": "run_backtest(symbol='AAPL', strategy_type='sma_cross')", "ml_strategy": "run_backtest(symbol='TSLA', strategy_type='online_learning')", "get_help": "Use prompts like 'backtest_strategy_guide' for detailed guides", }, } @mcp.tool(name="list_all_strategies") async def list_all_strategies() -> list[dict[str, Any]]: """ List all available backtesting strategies with their parameters. Returns detailed information about each strategy including: - Strategy name and description - Required and optional parameters - Default parameter values - Example usage """ strategies = [] # Return as array # Traditional strategies strategies.extend( [ { "name": "sma_cross", "description": "Simple Moving Average Crossover", "parameters": { "fast_period": { "type": "int", "default": 10, "description": "Fast MA period", }, "slow_period": { "type": "int", "default": 20, "description": "Slow MA period", }, }, "example": "run_backtest(symbol='AAPL', strategy_type='sma_cross', fast_period=10, slow_period=20)", }, { "name": "rsi", "description": "RSI Mean Reversion", "parameters": { "period": { "type": "int", "default": 14, "description": "RSI calculation period", }, "oversold": { "type": "int", "default": 30, "description": "Oversold threshold", }, "overbought": { "type": "int", "default": 70, "description": "Overbought threshold", }, }, "example": "run_backtest(symbol='MSFT', strategy_type='rsi', period=14)", }, { "name": "macd", "description": "MACD Signal Line Crossover", "parameters": { "fast_period": { "type": "int", "default": 12, "description": "Fast EMA period", }, "slow_period": { "type": "int", "default": 26, "description": "Slow EMA period", }, "signal_period": { "type": "int", "default": 9, "description": "Signal line period", }, }, "example": "run_backtest(symbol='GOOGL', strategy_type='macd')", }, { "name": "bollinger", "description": "Bollinger Bands Mean Reversion", "parameters": { "period": { "type": "int", "default": 20, "description": "BB calculation period", }, "std_dev": { "type": "float", "default": 2, "description": "Standard deviations", }, }, "example": "run_backtest(symbol='AMZN', strategy_type='bollinger')", }, { "name": "momentum", "description": "Momentum Trading Strategy", "parameters": { "period": { "type": "int", "default": 10, "description": "Momentum period", }, "threshold": { "type": "float", "default": 0.02, "description": "Entry threshold", }, }, "example": "run_backtest(symbol='NVDA', strategy_type='momentum')", }, { "name": "ema_cross", "description": "Exponential Moving Average Crossover", "parameters": { "fast_period": { "type": "int", "default": 12, "description": "Fast EMA period", }, "slow_period": { "type": "int", "default": 26, "description": "Slow EMA period", }, }, "example": "run_backtest(symbol='META', strategy_type='ema_cross')", }, { "name": "mean_reversion", "description": "Statistical Mean Reversion", "parameters": { "lookback": { "type": "int", "default": 20, "description": "Lookback period", }, "entry_z": { "type": "float", "default": -2, "description": "Entry z-score", }, "exit_z": { "type": "float", "default": 0, "description": "Exit z-score", }, }, "example": "run_backtest(symbol='SPY', strategy_type='mean_reversion')", }, { "name": "breakout", "description": "Channel Breakout Strategy", "parameters": { "lookback": { "type": "int", "default": 20, "description": "Channel period", }, "breakout_factor": { "type": "float", "default": 1.5, "description": "Breakout multiplier", }, }, "example": "run_backtest(symbol='QQQ', strategy_type='breakout')", }, { "name": "volume_momentum", "description": "Volume-Weighted Momentum", "parameters": { "period": { "type": "int", "default": 10, "description": "Momentum period", }, "volume_factor": { "type": "float", "default": 1.5, "description": "Volume multiplier", }, }, "example": "run_backtest(symbol='TSLA', strategy_type='volume_momentum')", }, ] ) # ML strategies strategies.extend( [ { "name": "ml_predictor", "description": "Machine Learning predictor using Random Forest", "parameters": { "model_type": { "type": "str", "default": "random_forest", "description": "ML model type", }, "n_estimators": { "type": "int", "default": 100, "description": "Number of trees", }, "max_depth": { "type": "int", "default": None, "description": "Max tree depth", }, }, "example": "run_ml_strategy_backtest(symbol='AAPL', strategy_type='ml_predictor', model_type='random_forest')", }, { "name": "online_learning", "description": "Online learning adaptive strategy (alias for adaptive)", "parameters": { "learning_rate": { "type": "float", "default": 0.01, "description": "Adaptation rate", }, "adaptation_method": { "type": "str", "default": "gradient", "description": "Method for adaptation", }, }, "example": "run_ml_strategy_backtest(symbol='AAPL', strategy_type='online_learning')", }, { "name": "regime_aware", "description": "Market regime detection and adaptation", "parameters": { "regime_window": { "type": "int", "default": 50, "description": "Regime detection window", }, "threshold": { "type": "float", "default": 0.02, "description": "Regime change threshold", }, }, "example": "run_backtest(symbol='SPY', strategy_type='regime_aware')", }, { "name": "ensemble", "description": "Ensemble voting with multiple strategies", "parameters": { "fast_period": { "type": "int", "default": 10, "description": "Fast MA for ensemble", }, "slow_period": { "type": "int", "default": 20, "description": "Slow MA for ensemble", }, "rsi_period": { "type": "int", "default": 14, "description": "RSI period for ensemble", }, }, "example": "run_ml_strategy_backtest(symbol='MSFT', strategy_type='ensemble')", }, { "name": "adaptive", "description": "Adaptive strategy that adjusts based on performance", "parameters": { "learning_rate": { "type": "float", "default": 0.01, "description": "How quickly to adapt", }, "adaptation_method": { "type": "str", "default": "gradient", "description": "Method for adaptation", }, }, "example": "run_ml_strategy_backtest(symbol='GOOGL', strategy_type='adaptive')", }, ] ) return strategies # Return array @mcp.tool(name="get_strategy_help") async def get_strategy_help(strategy_type: str) -> dict[str, Any]: """ Get detailed help for a specific strategy. Args: strategy_type: Name of the strategy (e.g., 'sma_cross', 'online_learning') Returns: Detailed information about the strategy including theory, parameters, and best practices. """ strategy_help = { "sma_cross": { "name": "Simple Moving Average Crossover", "theory": "Generates buy signals when fast SMA crosses above slow SMA, sell when opposite occurs", "best_for": "Trending markets with clear directional moves", "parameters": { "fast_period": "Typically 10-20 days for short-term trends", "slow_period": "Typically 20-50 days for medium-term trends", }, "tips": [ "Works best in trending markets", "Consider adding volume confirmation", "Use wider periods for less noise", ], }, "ml_predictor": { "name": "Machine Learning Predictor", "theory": "Uses Random Forest or other ML models to predict price movements", "best_for": "Complex markets with multiple factors", "parameters": { "model_type": "Type of ML model (random_forest)", "n_estimators": "Number of trees in forest (50-200)", "max_depth": "Maximum tree depth (None or 5-20)", }, "tips": [ "More estimators for better accuracy but slower", "Limit depth to prevent overfitting", "Requires sufficient historical data", ], }, "online_learning": { "name": "Online Learning Strategy", "theory": "Continuously adapts strategy parameters based on recent performance", "best_for": "Dynamic markets with changing patterns", "parameters": { "learning_rate": "How quickly to adapt (0.001-0.1)", "adaptation_method": "Method for adaptation (gradient, bayesian)", }, "tips": [ "Lower learning rates for stable adaptation", "Works well in volatile markets", "This is an alias for the adaptive strategy", ], }, "adaptive": { "name": "Adaptive Strategy", "theory": "Dynamically adjusts strategy parameters based on performance", "best_for": "Markets with changing characteristics", "parameters": { "learning_rate": "How quickly to adapt (0.001-0.1)", "adaptation_method": "Method for adaptation (gradient, bayesian)", }, "tips": [ "Start with lower learning rates", "Monitor for overfitting", "Works best with stable base strategy", ], }, "ensemble": { "name": "Strategy Ensemble", "theory": "Combines multiple strategies with weighted voting", "best_for": "Risk reduction through diversification", "parameters": { "base_strategies": "List of strategies to combine", "weighting_method": "How to weight strategies (equal, performance, volatility)", }, "tips": [ "Combine uncorrelated strategies", "Performance weighting adapts to market", "More strategies reduce single-point failure", ], }, "regime_aware": { "name": "Market Regime Detection Strategy", "theory": "Identifies market regimes (trending vs ranging) and adapts strategy accordingly", "best_for": "Markets that alternate between trending and sideways movement", "parameters": { "regime_window": "Period for regime detection (30-100 days)", "threshold": "Sensitivity to regime changes (0.01-0.05)", }, "tips": [ "Longer windows for major regime shifts", "Lower thresholds for more sensitive detection", "Combines well with other indicators", ], }, } if strategy_type in strategy_help: return strategy_help[strategy_type] else: return { "error": f"Strategy '{strategy_type}' not found", "available_strategies": list(strategy_help.keys()), "tip": "Use list_all_strategies() to see all available strategies", }