stock-predictor-pro

What problem does it solve?

This Skill helps you predict A-share stock upside potential using large-scale historical data and machine-learning ensembles, so you can quickly筛选出更值得关注的股票并生成每日推荐。

Core Features & Use Cases

• Ensemble stock prediction: Uses an integrated ensemble approach (LightGBM + XGBoost + Random Forest) with feature engineering and a probability threshold to rank candidate stocks.
• Threshold optimization & validation: Supports systematic threshold tuning and backtesting/verification to reduce noisy screening.
• Daily Top10 recommendation generation: Reads latest market features from a SQLite database/market data pipeline, then outputs Top10 lists for today and tomorrow.

Use Case: You want a repeatable workflow to transform daily market updates into a prioritized Top10 watchlist based on an ML-driven probability score (e.g., threshold=0.65) without manually running training or evaluation each day.

Quick Start

Ask the system: run the stock-predictor-pro skill to generate today and tomorrow Top10 recommendations using the configured threshold and the local SQLite database.