Oracle3
Automate paper-trading and arbitrage detection across prediction markets by checking no-arbitrage constraints and managing risk.
Oracle3 is an open-source paper-trading engine and MCP server designed for prediction markets like Kalshi and Polymarket. It maps logical relationships between event contracts, assesses whether quoted prices violate the axioms of probability after accounting for each venue's fees, and paper-trades profitable baskets under predefined risk limits. This tool helps users discover and exploit pricing inefficiencies, providing comprehensive market analysis and risk-managed simulated trading capabilities.
Key Features
01Multi-venue prediction market support (Kalshi, Polymarket)
02Automated no-arbitrage constraint checking across related contracts
03Comprehensive fee schedule integration for accurate net edge calculation
04Paper trading engine with pre-trade risk management and kill switch
05Multi-interface access: MCP server, JSON CLI, and Python API
06258 GitHub stars
Use Cases
01Integrating AI agents with prediction market data for automated analysis and paper-trading.
02Conducting quantitative research on pricing inefficiencies and fee structures in prediction markets.
03Developing and backtesting algorithmic trading strategies for prediction markets.