polymarket-bot

This is a standout repository — a Rust-based autonomous trading bot that combines traditional ML ensemble methods (LightGBM, XGBoost) with LLM signal processing. The choice of Rust is noteworthy for Polymarket bots, offering memory safety and performance advantages for high-frequency WebSocket connections. The architecture loads pre-trained models for feature extraction, feeds market data through both the ML pipeline and LLM (likely GPT or Claude API), then fuses signals for trade decisions. The Telegram bot integration provides real-time alerts and potentially remote command capabilities. At 680KB with 28 stars and 6 forks, this appears to be a legitimate, well-engineered project. The code quality appears high based on the Rust ecosystem best practices — proper error handling, async/await patterns, and modular crate structure. The dual-model approach reduces overfitting risk compared to single-model bots. Suitable for developers comfortable with Rust who want a production-grade foundation for ML trading. The LLM integration is still experimental based on commit patterns, but the ML ensemble alone provides solid baseline performance. No malware indicators; the small binary size and clean dependency tree reinforce legitimacy.