What it is
An MCP server exposing ten Jev judgment tools for agents: claim verification, content screening, candidate ranking, classification, comparison, extraction, diff review, and gating.
How it uses Jev
Each tool sends a state (claims, text, candidates, items, passages, or a diff) plus typed questions to Jev and returns calibrated probabilities, verdicts, and confidence scores. The calling agent uses these results to decide whether to accept, review, or block.
Primitives:choicescorenoul
Technique worth stealing
Use cheap, fast typed judgments for mechanical checks agents would otherwise skip.
Try it
npx -y @jkudish/jev-mcp
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-centricyes0.91
- Shows a System One patternno0.32
- Handles uncertaintyno0.05
- Measuredno0.07
- Runnableyes0.93
- Worth recommendingno0.37
- Model replicano0.05
- Problem scopescore on a 0–2 scale1.62
- About Jevyes0.98
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.