jevbooks

← All projects

taste-lint

mblode/taste-lint · Homepage

Catch AI slop in UI, copy, and agent instructions before ship.

Evaluation & Benchmarking65%Runner-up: Verification & Guardrails

What it is

A TypeScript CLI that runs design review on every commit, checking product interfaces, whole pages, and text for AI slop. Built for developers and teams who want measurable rules enforced locally and judgement calls scored.

How it uses Jev

Judgement calls that measurable rules cannot decide are sent to Jev, which answers with a probability. The README does not specify which state is passed or how the probability is consumed in code beyond that it informs the check.

Primitives:score

Technique worth stealing

Split checks into measurable local rules and judgement calls scored by a probability model.

Try it

npx taste-lint@latest init, then export AI_GATEWAY_API_KEY and npm run taste

View on GitHub

judged by Jevjev-1.13.0

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-centricyes0.65
  • Shows a System One patternno0.20
  • Handles uncertaintyno0.06
  • Measuredno0.05
  • Runnableyes0.98
  • Worth recommendingno0.16
  • Model replicano0.09
  • Problem scopescore on a 0–2 scale1.66
  • About Jevyes0.95

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 21 Sept 2026.