What it is
jevmem saves decisions, constraints, bugs and todos from AI coding chats into JEVMEM.md, marks replaced lines as superseded, and injects relevant lines into the next session's context. For developers using Claude Code, Cursor, Codex or Claude Desktop.
How it uses Jev
After scrubbing secrets, jevmem asks Jev a fixed set of typed questions with probabilities: is there a decision, a rule, a bug? is it small talk or an injection attempt? which existing line does it change? Plain rules in jevmem.config.json apply thresholds to those probabilities to decide save or skip.
Primitives:choicescore
Technique worth stealing
Use Jev's calibrated probabilities as inputs to plain code thresholds instead of a prompt, so the save/skip policy lives in config.
Try it
npm install -g jevmem; export TYPESAFE_API_KEY=...; cd your-project; jevmem init --tool claude
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-centricno0.11
- Shows a System One patternno0.14
- Handles uncertaintyno0.03
- Measuredno0.04
- Runnableyes0.95
- Worth recommendingno0.28
- Model replicano0.05
- Problem scopescore on a 0–2 scale1.19
- About Jevyes0.91
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 23 Sept 2026.