What it is
A Python router for a DJ chatbot that classifies free-form /play requests before an LLM handles them, producing a structured hint for the LLM.
How it uses Jev
Jev classifies each /play request in one call using 9 typed questions on the raw request text. It returns route probabilities, noul answers, and a score with confidence. These are turned into a hint (route, alternatives, extract, modifiers, energy) that guides the LLM or triggers direct actions.
Primitives:choicescorenoul
Technique worth stealing
Use Jev as a cheap pre-classifier to route requests and generate hints for an LLM.
Try it
./play.sh yellow from coldplay
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-centricyes0.83
- Shows a System One patternno0.24
- Handles uncertaintyno0.14
- Measuredunclear0.49
- Runnableno0.38
- Worth recommendingunclear0.54
- Model replicano0.03
- Problem scopescore on a 0–2 scale1.09
- About Jevyes0.99
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 23 Sept 2026.