jevbooks

← All projects

jev-study

wjdjdakf17/jev-study

Korean study notes and TypeScript mockups for Jev's typed decisions

What it is

A personal Korean-language study repo on Jev, TypeSafe AI's "System One Model". It turns Jev's published concepts into TypeScript types, a deterministic mock client, a routing workflow, and calibration math, aimed at readers who want to understand the ideas before the API is public.

How it uses Jev

The repo does not call the real Jev API. Instead src/client.ts defines a DecisionClient interface with a deterministic mock implementation, and src/workflow.ts shows confidence-gated routing: a decision's calibrated probability is compared against a threshold, and low-confidence tickets escalate to human review.

Primitives:choicescorenoul

Technique worth stealing

Confidence-gated routing: let the model judge, let an if-statement on the returned probability decide whether to act or escalate.

Try it

git clone the repo, then npm install, npm test (26 tests), npm run example.

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-centricyes0.95
  • Shows a System One patternyes0.92
  • Handles uncertaintyno0.24
  • Measuredno0.08
  • Runnableno0.12
  • Worth recommendingno0.39
  • Model replicayes0.79
  • Problem scopescore on a 0–2 scale1.08
  • About Jevyes0.98

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.