What it is
A router extension for an existing local Codex Router installation. It picks a model and thinking effort for each Codex model call, including continuations after tools, aiming for sufficient capability without unnecessary quota use.
How it uses Jev
Jev classifies each turn using the current request, recent assistant intent, and available tool result, then chooses one of 15 model/effort pairs (Luna, Sol, Astra × low/medium/high/xhigh/max) in a single Choice question. The decision sets model, reasoning.effort, and service_tier for the call.
Primitives:choice
Technique worth stealing
Fail-open routing: any Jev error keeps the turn alive with a safe fallback route, and a sentinel file bypasses Jev instantly.
Try it
Run server/report_routing.py to see routing and savings; install via AGENTS.md for AI agents.
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 patternyes0.77
- Handles uncertaintyno0.05
- Measuredyes0.65
- Runnableyes0.82
- Worth recommendingunclear0.48
- Model replicano0.08
- Problem scopescore on a 0–2 scale1.08
- About Jevyes0.98
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 21 Sept 2026.