User repository on GitHub
- Forks
- 0
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.Homepage
CommunityJev Demos & Experiments
Jev-test
A TypeScript CLI and Fastify service that puts three questions about an occupation to Jev in one request - layoff risk, trajectory and human accountability - and derives a resilience score in app code.
- Stars
- 0
- Language
- TypeScript
- Published by
- Community
- Author
- WeSecureYou
- Last verified
- Oct 2026
About this project
A TypeScript/Node.js CLI and REST API that uses TypeSafe AI's Jev System One model to evaluate an occupation's exposure to AI-driven layoffs, likely future trajectory, need for human accountability, and overall resilience. The project works immediately without credentials. When JEVAPIKEY is absent, it logs a warning and uses deterministic, occupation-aware mock data with the same validated shape as a live Jev response. Optional: a TypeSafe AI API key for live analysis.
Quoted from the project's README. Read the full README.
Used inJev for Software & DevTools
- Languages
- TypeScript
Highlights
- One POST to /v1/systemone on jev-latest carries a five-level Score, a three-way Choice and a Noul question.
- Without JEV_API_KEY the CLI and server use deterministic occupation-aware mock data with the same validated shape.
- Output reports layoff risk, future trajectory, trajectory confidence, human accountability and a derived resilience score.
- It ships a browser UI at localhost:3000 that calls the REST API, plus npm scripts for dev, server, cli and typecheck.
Watch out
The resilience score is not a model output: the README describes it as a transparent application-level calculation combining inverse layoff risk, human accountability and a small trajectory adjustment, without publishing the weights. The mock fallback is easy to miss in a demo - the README says a missing or blank variable only logs a warning, though the response does carry source: mock and model: jev-latest-mock. The server listens on 0.0.0.0:3000 by default and the README documents HOST and PORT overrides but no authentication. The five-level layoff_risk Score is normalised to 0..1 in application code. The README says results are forecasts, not employment or financial advice, and gives no evaluation set, accuracy figure or question wording. Copy reviewed against the repository README on 2026-10-04, not from running it.
Questions about this entry
Short answers drawn from the linked project or page.
You may be reading mock numbers, and the response tells you. When JEV_API_KEY is unset or blank the CLI and server log a warning and fall back to deterministic, occupation-aware mock data validated into the same shape as a live Jev response. The example payload carries "source": "mock" and "model": "jev-latest-mock", so check that field before treating a forecast as the model's judgment.
By default, yes: the README states it listens on 0.0.0.0:3000 and that HOST and PORT override it. There is no authentication described anywhere in the README, and the browser UI is served from the same host, so on a shared or untrusted network you would want to set HOST to loopback yourself. The Vercel path is documented separately: the Fastify entrypoint in src/index.ts is detected without extra config.
The code. Only three questions go to the model: layoff_risk as a five-level Score normalised to 0..1 in application code, future_trajectory as a Choice among high_layoff_risk, transformed_hybrid and ruling_trend, and human_accountability as a Noul. The resilience score is an application-level combination of inverse layoff risk, human accountability and a small trajectory adjustment; the README says it is transparent but does not publish the exact weights.
The README does not say. There is no evaluation set, no accuracy number, no comparison against any other model and no question wording in it - only the field names and types. Treat it as a working integration shape rather than a validated instrument, which is consistent with the README calling the output a forecast rather than employment or financial advice.
Run npm run cli -- --title "Data Entry Clerk", adding --desc with role context when you have it, or invoke npx tsx src/cli/index.ts directly. The table reports layoff risk, future trajectory, trajectory confidence, human-accountability probability and the derived resilience score.
Start the server with npm run server, or npm run dev to restart on TypeScript changes, and POST to /api/analyze with a title and description. There is a /health check, and the browser UI at localhost:3000 calls the same API from the report panel.
Run npx vercel deploy --prod. Vercel detects the Fastify entrypoint in src/index.ts without extra configuration; add JEV_API_KEY in the project's environment variables for live analysis, and without it the deployed API uses the typed mock fallback.
More Jev resources like this
23GitHub stars
Chrome extension that sends each X post to Jev with six typed questions in one call near the viewport, then renders a behavior chip under the post plus a session cost and latency panel.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
oso95/x-scannerTypeScript
oso95/x-scannerTypeScript2GitHub stars
A Magic 8 Ball that picks with Jev instead of at random. Hold, shake, release, and Jev chooses from the twenty classic answers and returns a probability for each one, shown under the ball.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
mikecann/magic-jev-ballTypeScript
mikecann/magic-jev-ballTypeScript0GitHub stars
JevPlay is a playground for TypeSafe Jev: a freeform state block plus a builder for custom Choice, Score and Noul questions, showing live probability distributions as you write them. Setup is npm i and npm run dev.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
ndolinschi/jevplayTypeScript
ndolinschi/jevplayTypeScript0GitHub stars
Chrome extension that re-ranks Google results with TypeSafe Jev: five atomic questions per result, a weighted score, sales pages flagged and hidden, and a button to restore Google's order.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
tylergibbs1/siftTypeScript
tylergibbs1/siftTypeScript46GitHub stars
Open 150M encoder that reproduces the typed noul, choice, and score interface in one forward pass, trainable on a free GPU and runnable locally for nothing.
Jev Demos & Experiments#community#jevbest#catalog-import
98GitHub stars
Self-hosted Docker service and web UI that packages Laya structured-decision inference behind a compatible System One endpoint.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
1Panel-dev/laya-serverPython
1Panel-dev/laya-serverPythonFrom the Jev guides
Original write-ups on what this entry is for.
A game tick is a state plus a legal action list. How Clash Royale, Minecraft, and the earlier Mario, Pokémon, StarCraft, and drone loops ask Jev one step at a time.
Jev Demos & Experiments#gaming#agents#control
Read article
From the Jev community
Posts from builders shipping with Jev right now.