What problem does it solve? Writing and running quantitative trading strategies requires deep knowledge of the Qlib framework, correct stock code formats, backtest engine configuration, and containerized execution environments. This Skill lets an AI assistant generate valid Qlib strategy code from natural language, execute it in an isolated Docker container, stream backtest results, and persist strategies to the cloud. ## Core Features & Use Cases - AI Strategy Generation: Produces module-type strategies with STRATEGY_CONFIG or get_strategy_config() compatible with the backtest center, supporting 11 strategy classes such as TopK, weight-based, long-short, stop-loss, and sector rotation. - Containerized Execution & Streaming Logs: Starts strategy runs or backtests via /api/v1/ai-ide/execute endpoints, monitors SSE log streams for results like annual return and Sharpe ratio, and supports stop and syntax-check operations. - Cloud Strategy Persistence: Saves generated strategies with parameters and tags through the /api/v1/strategies API for later reuse and iteration. - Use Case: A user asks the AI to "write a low-valuation strategy selecting stocks with PE under 15 and ROE above 10, then backtest it on the CN market" — the Skill generates the config, saves it, runs the backtest with the default model, and streams the results. ## Quick Start Ask the AI to write a Qlib TopK strategy for the A-share market with your stock selection conditions, save it to the cloud, run the backtest, and show the annual return and Sharpe ratio.
