What it is
A local, free, Jev-compatible System 1 AI stack for developers and Salesforce teams: a backend API, Salesforce LWC demo, and a Chrome extension that injects document.modelContext into any web page.
How it uses Jev
The backend's POST /v1/systemone accepts Jev-format requests with a state string and typed questions, then returns calibrated Noul, Choice, and Score answers with zero output tokens. Salesforce LWCs call it via Apex or run Transformers.js in a Web Worker; the Chrome extension exposes registered tools through document.modelContext.
Primitives:noulchoicescore
Technique worth stealing
Run a local NLI model (Transformers.js or Cloudflare Workers AI) behind a Jev-compatible API to get calibrated Noul/Choice/Score answers with zero output tokens.
Try it
cd apps/backend && bun install && bun run dev, or npx wrangler deploy for Cloudflare Workers.
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-centricyes0.88
- Shows a System One patternno0.26
- Handles uncertaintyno0.05
- Measuredno0.39
- Runnableyes0.95
- Worth recommendingno0.21
- Model replicayes0.84
- Problem scopescore on a 0–2 scale1.03
- About Jevyes0.97
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.