What it is
An experimental Python framework for forecasting short-term Hong Kong stock price direction. It targets researchers who want a disciplined reasoning layer over structured market data, not a proven trading model.
How it uses Jev
Jev answers one choice question per requested trading day plus a separate 30-calendar-day endpoint question, given a past-only state of the target stock, HSI, and HSTECH. The program validates the three-class probabilities and saves JSON or an HTML report.
Primitives:choice
Technique worth stealing
Keep data retrieval, feature calculation, leakage checks, and scoring in Python; let Jev make only the classification choice.
Try it
pip install -r requirements.txt, set TYPESAFE_API_KEY in .env, then run jev_hk_predict.py predict --symbol 01810 --days 3.
Evidence
Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.
- Jev-centricyes0.95
- Shows a System One patternunclear0.49
- Handles uncertaintyno0.06
- Measuredno0.08
- Runnableunclear0.53
- Worth recommendingunclear0.40
- Model replicano0.11
- Problem scopescore on a 0–2 scale0.65
- About Jevyes0.98
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.