LLM-powered trading agents that turn plain natural language into a five-pillar strategy: Trend, Mean-Reversion, Momentum, Volume, and Risk. Each strategy is hosted, self-evolving, configurable through 30+ tunable parameters, and bit-exact between backtest and live execution. Built for simulated Hyperliquid perpetuals.

Provide your stock strategy idea and let your agent implement and improve it.

Local first paper trading and backtesting lab with historical OHLCV replay, simulated order execution, SQLite storage, and a small API.

[🚧] - Making an event streaming trading engine powered by AI decision