This MCP server specializes in market microstructure analysis and manipulation detection, orchestrating 14 financial data actors in parallel to deliver production-grade econometric methods. It's designed for quantitative researchers, compliance teams, and algorithmic traders, offering advanced tools like Hawkes processes, BOCPD, spectral transfer entropy, Granger causality, and game-theoretic surveillance optimization without the burden of infrastructure management. Users can connect any MCP-compatible AI client to analyze order book dynamics, detect various manipulation types, measure cross-asset information flow, decompose bid-ask spreads, identify abnormal insider flows, classify market regimes, and optimize surveillance resource allocation, all backed by live data from diverse financial sources.
主要功能
01Bayesian Online Changepoint Detection (BOCPD) for manipulation patterns
02Queue-reactive Hawkes process estimation
03Spectral transfer entropy for directed information flow across assets
04Integration with 14 parallel data sources including Finnhub, SEC EDGAR, and CoinGecko
050 GitHub stars
06Extensive-form game theory for surveillance budget optimization
