crypto-quant-signal-mcp

Generates trading signals by analyzing cryptocurrency market data through quantitative models and technical indicators. The MCP server exposes signal generation as a tool that Claude and other LLM clients can invoke, passing market parameters and receiving structured signal outputs (buy/sell/hold recommendations with confidence scores). Integrates with crypto data providers to fetch OHLCV data, compute indicators (RSI, MACD, Bollinger Bands, etc.), and apply statistical models to produce actionable trading signals.

Computes technical indicators (RSI, MACD, Bollinger Bands, moving averages, etc.) across multiple timeframes simultaneously and aggregates results into a unified signal consensus. The server accepts a cryptocurrency symbol and returns indicator values for 1m, 5m, 15m, 1h, 4h, 1d timeframes in a single call, enabling multi-timeframe analysis patterns where short-term and long-term trends are evaluated together. Aggregation logic may apply voting or weighting schemes to produce a composite signal strength.

Implements comprehensive error handling and input validation for MCP protocol interactions, including parameter schema validation, graceful error responses, and detailed error messages. The server validates all inputs against JSON Schema definitions and returns structured error responses with actionable guidance.

Validates signal generation logic against historical cryptocurrency price data by replaying signals across past market conditions and computing performance metrics (win rate, profit factor, max drawdown, Sharpe ratio). The server accepts a signal strategy definition (indicator parameters, entry/exit rules) and a date range, then simulates trades on historical OHLCV data and returns detailed backtest results. Enables rapid iteration on signal parameters without live trading risk.

Connects to cryptocurrency exchange WebSocket APIs (Binance, Kraken, Coinbase, etc.) to stream real-time OHLCV data, order book updates, and trade ticks. The MCP server maintains persistent WebSocket connections and exposes current market state as queryable tools, allowing Claude to request the latest price, volume, bid-ask spread, or order book depth for any crypto asset. Handles reconnection logic, data buffering, and timestamp synchronization automatically.

Analyzes price correlations and technical indicator divergences across multiple cryptocurrency assets to identify trading opportunities. Computes rolling correlations between asset pairs (e.g., BTC vs ETH, BTC vs altcoins), detects when correlated assets diverge from expected relationships, and identifies bullish/bearish divergences (price makes new high but indicator does not). Returns structured divergence alerts with asset pairs, divergence type, and confidence scores.

Analyzes historical volatility patterns to detect current market regime (low volatility, high volatility, volatility expansion, volatility contraction) and forecasts near-term volatility using models like GARCH or exponential weighted moving average (EWMA). Returns current volatility regime classification, volatility percentile (current vol vs historical range), and volatility forecast for next N candles. Enables regime-aware signal filtering and position sizing.

Allows users to define custom alert rules (e.g., 'alert when BTC price crosses above 50-day MA', 'alert when RSI > 70 on 1h timeframe', 'alert when correlation between BTC and ETH drops below 0.7') and automatically monitors market conditions to trigger notifications. The MCP server maintains rule state, evaluates conditions against real-time or periodic market data, and exposes tools to create, update, delete, and list active alerts. Supports multiple notification channels (webhook, email, Slack, Discord, etc.).

Tracks cryptocurrency portfolio holdings and computes performance metrics (total return, daily/monthly return, Sharpe ratio, max drawdown, win rate) with attribution analysis showing which assets/trades contributed most to gains or losses. Accepts portfolio composition (asset holdings, entry prices, quantities) and compares against market benchmarks (BTC, ETH, market cap-weighted index). Returns detailed performance breakdown by asset and time period.

Integrates external data sources (social media sentiment, on-chain metrics, funding rates, open interest) into the signal generation pipeline. Fetches sentiment scores from crypto-specific sources (Twitter/X sentiment, Reddit mentions, news sentiment), on-chain metrics (whale transactions, exchange inflows/outflows, active addresses), and derivatives data (funding rates, open interest, liquidation levels). Combines these signals with technical indicators to produce more robust trading signals.

Analyzes signal performance and attributes returns to specific signal components (trend component, mean-reversion component, volatility component, etc.), providing detailed performance analytics and component contribution metrics. The server decomposes signal performance and identifies which signal components drive profitability.