What it is
A router that maps an AI agent's visualization request to one of six rendering branches (scientific plots, architecture diagrams, slides, AI-generated visuals, materials computation, chart MCP), returning a target tool and output format.
How it uses Jev
The README describes Jev as a System-1 decision engine that classifies the drawing query in under 300ms and returns a branch, target tool, rendering format, confidence and fast-path flag. It does not state which Jev primitives (Choice/Score/Noul) are used or how the state is constructed.
Technique worth stealing
Fast-path routing: if confidence >= 0.85, invoke the specialized tool directly instead of loading the full router table.
Try it
git clone the repo, run ./install.sh, set OPENROUTER_API_KEY, then python3 scripts/fast_route.py -q "..." --template
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 patternunclear0.57
- Handles uncertaintyno0.09
- Measuredno0.09
- Runnableyes0.82
- Worth recommendingno0.29
- Model replicano0.11
- Problem scopescore on a 0–2 scale1.49
- About Jevyes0.93
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.