What it is
A semantic code search CLI for developers. It asks a yes/no question of every function in a codebase and ranks the resulting probabilities, so you can find code matching a described behavior rather than a text pattern.
How it uses Jev
Jev judges each function against the user's question in its own question, returning a probability that the answer is yes. Scores are ranked, and results within ±0.10 of the threshold are asked again and averaged. The CLI uses the scores to order and filter functions.
Primitives:choice
Technique worth stealing
One Jev question per function, never a shared list, because positional lookup degrades past ~16 items.
Try it
pip install every-cli; export TYPESAFE_API_KEY=...; every "<question>" <path>
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-centricyes0.87
- Shows a System One patternunclear0.52
- Handles uncertaintyno0.31
- Measuredyes0.97
- Runnableyes0.93
- Worth recommendingno0.28
- Model replicano0.16
- Problem scopescore on a 0–2 scale1.75
- About Jevyes0.98
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.