jevbooks

← All projects

huncho

edgardcham/huncho

Decisions as code on System One models: typed questions, thresholds with hysteresis, nested decisions, journal, calibration

What it is

huncho is a TypeScript SDK for developers to define decisions as code. It asks typed questions of a decision model, applies thresholds with hysteresis, nests decisions, journals every one, replays policy changes without inference, and calibrates against outcomes.

How it uses Jev

huncho uses Jev via providers (direct, OpenRouter, Vercel AI Gateway). It asks typed questions (Noul, Choice, Score) of Jev, receives probabilities, and applies thresholds with hysteresis to produce named decisions. The result is used in code as decision.outcome, decision.via, and decision.previous.

Primitives:noulchoicescore

Technique worth stealing

Hysteresis thresholds with enter/exit values to prevent flapping, per key.

Try it

npm i huncho, then run examples with TYPESAFE_API_KEY set: node --env-file-if-exists=.env.local dist/examples/support-route.js

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-centricno0.31
  • Shows a System One patternyes0.92
  • Handles uncertaintyno0.33
  • Measuredno0.07
  • Runnableyes0.96
  • Worth recommendingunclear0.49
  • Model replicano0.04
  • Problem scopescore on a 0–2 scale1.13
  • About Jevyes0.98

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