What it is
A Claude Code plugin that adds TypeSafe Jev semantic judgments to Bash, Write, and Edit tool calls, warning or asking before potentially destructive, exfiltrating, or out-of-scope actions.
How it uses Jev
Pre-tool gate sends four questions (destructive, exfiltration, beyondScope, impact) on bounded tool input and user request; output gate sends leaks_secret and failure_class on Bash output. Local threshold composition turns typed answers into warnings or permissionDecision ask.
Primitives:choicescorenoul
Technique worth stealing
Shadow mode by default; enforce mode returns native permissionDecision ask, never allow from TypeSafe confidence.
Try it
/plugin marketplace add dr-dimitru/claude-plugins-marketplace; /plugin install claude-jev@dr-dimitru-claude-tools --scope user
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-centricyes0.78
- Shows a System One patternno0.17
- Handles uncertaintyno0.12
- Measuredno0.03
- Runnableyes0.98
- Worth recommendingno0.20
- Model replicano0.04
- Problem scopescore on a 0–2 scale1.12
- About Jevyes0.97
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 21 Sept 2026.