What it is
Jev Review is a local-first MCP plugin that gives Claude Code, Codex, Cursor, and OpenCode structured software-quality scores while they work, across correctness, complexity, changeability, modularity, tests, security, and other dimensions.
How it uses Jev
The agent calls the single MCP tool `jev_review` with a task, focused diff, optional files, repository context, and a previous evaluation. Jev returns typed Score, Choice, and Noul decisions, which Jev Review converts into metric scores, confidence, rubric hints, and deltas; the agent diagnoses and changes the code.
Primitives:choicescorenoul
Technique worth stealing
Repeated scalar feedback loop: baseline score, agent hypothesis, smallest justified improvement, rescore with previousEvaluation to see deltas.
Try it
export JEV_API_KEY="your-key"; npx plugins add NiazMorshed2007/jev-review; restart your client.
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-centricyes0.77
- Shows a System One patternno0.22
- Handles uncertaintyno0.06
- Measuredno0.05
- Runnableyes0.96
- Worth recommendingno0.34
- Model replicano0.06
- Problem scopescore on a 0–2 scale1.57
- About Jevyes0.98
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.