build-trading-strategies

To generate crypto trading strategy code using AI, define a clear strategy concept and verify data availability, then use the create_strategy tool to output a production-ready Jesse Python class with entry, exit, sizing, and risk management logic. This provides a complete foundation for backtesting and live deployment.

Creating a prediction market strategy for Polymarket uses the create_prediction_market_strategy tool to generate YES/NO strategy code. This outputs a Python class built for prediction market frameworks, enabling automated trading logic specifically tailored for binary outcome markets.

Yes, generated strategy code includes risk management and position sizing hooks. The output Jesse-based Python class features should_long, should_short, go_long, and go_short methods, plus optional on_open_position and update_position hooks for executing dynamic risk controls during live trading.

The best way to validate AI-generated crypto strategies is to run backtests using test-trading-strategies before deployment. This process evaluates the performance of the generated Jesse code against historical data, ensuring the entry, exit, and risk management logic behaves as expected.

Yes, you need to verify data before generating a Jesse trading strategy. Using browse-robonet-data to confirm indicator availability and symbols ensures the AI has accurate inputs to create a functional, production-ready strategy class ready for backtesting.

Yes, you can use this for both crypto and prediction market trading. The skill provides two distinct generation tools: create_strategy for standard crypto markets and create_prediction_market_strategy for Polymarket YES/NO strategies, both outputting production-ready Python code.