User repository on GitHub
Visually configure Jev questions, test them live, export ready-to-use code.
- Forks
- 1
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
CommunityJev Demos & Experiments
Jev Starter — YanfLIZi56
A browser console for Jev: build Noul, Choice and Score question sets, run them against your TypeSafe key, then export the same configuration as Python, JavaScript or cURL to drop into your own project.
- Stars
- 3
- Language
- Vue
- License
- MIT
- Published by
- Community
- Author
- YanfLIZi56
- Last verified
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is Jev Starter — hamakyo.
Works forJev for Classification
Used inJev for Software & DevTools
Highlights
- Runs with `npx @yanflizi56/jev-starter` on Node 18+ and Python 3.9+; the backend virtualenv is cached in your home directory and reused after the first launch.
- Export produces Python, JavaScript, cURL or JSON from the configuration you actually tested, rather than from a hand-written template.
- Configurations persist in browser local storage, so you can iterate on a question set and come back to it days later.
- An `examples/` directory holds community-contributed sets — chat emotion, resume screening, RAG relevance — importable straight into the console.
Watch out
It is a thin front end over the Jev API, not a local model: you paste a TypeSafe API key and every request the console makes is billed to you, so this is where you find out what a question set costs. Running from source means two processes — a Python backend and a Vite dev server with the API path proxied across — so console errors are not necessarily console bugs. The `examples/` table lists one file under one name and links to another, which reads like an unfinished rename. And the exports are only as good as the configuration: nothing here tells you whether your question set is a good one. 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 need the key. The console is a configuration UI and a code generator, not a local model — you enter a TypeSafe API key in the browser and it forwards your questions to the API, so every test run costs what the API costs. The README points anyone without a key to register first. What you save is engineering time: you find a question set that works in the UI instead of debugging request payloads inside your own code and inferring what the response shape should have been.
The configuration you tested. The loop is: define Noul, Choice and Score questions, optionally fill a state field, submit, read the result, then export as Python, JavaScript, cURL or JSON and copy it into your project. The emitted code reflects the question set you just ran, including the options and criteria you typed. One detail worth knowing: leaving state empty is a supported mode where the model answers from the instructions alone, so you can test a question set before deciding what context to feed it.
The first launch is slow and needs network access. The npx command creates a Python virtualenv and installs backend dependencies on first run, caching them under your home directory; later starts reuse the cache and open quickly. It needs Node 18+ and Python 3.9+. If the cached environment rots, a reset flag rebuilds it, and flags cover port selection and suppressing the browser. Running from source instead is two processes — backend plus Vite dev server with a proxied API path.
Worth starting from, with a caveat. There are three community-contributed question sets covering chat emotion analysis, resume pre-screening and RAG retrieval relevance, and importing one into the console is the fastest way to see how a well-formed set is structured. The visible rough edge: one table row is labelled with a file name that does not match the file it links to, which looks like an incomplete rename rather than two separate examples. Treat them as structural inspiration, not as validated prompts.
Yes, three ways. Share an example by building a question set, exporting it as Example JSON and opening a pull request into the examples directory, or paste the JSON into an issue and let a maintainer add it. You can also report bugs or request features through issues without touching any code.
In browser local storage. A question set you build persists automatically and is still there the next time you open the console, so iterating across days does not require re-entering it. There is no server-side account involved; the persistence is per browser.
Scenario plus author, for example chat-emotion-YanfLIZi56.json or resume-screening-YanfLIZi56.json. The flow is to export the tested set as an Example JSON and add it under examples/, either by pull request or by posting the file in an issue for a maintainer to place.
More Jev resources like this
46GitHub 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
941GitHub stars
Library that adds schema-guaranteed typed outputs to existing autoregressive LLMs without changing weights, served through SGLang or a hosted API.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
TypeLLM/TypeLLMPython
TypeLLM/TypeLLMPython98GitHub 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-serverPython39GitHub stars
Community Spring Boot starter that injects a JevClient into Spring MVC apps to ask typed noul, choice, and score questions against the TypeSafe API.
Jev Demos & Experiments#community#awesomejev#catalog-import
Community
danvega/jev-spring-boot-starterJava
danvega/jev-spring-boot-starterJava39GitHub stars
Mac menu-bar voice controller where Jev picks the next action from observed Accessibility controls and Swift executes it in an observe-decide-execute loop.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
ronadin2002/jev-cuaSwift
ronadin2002/jev-cuaSwift55GitHub stars
Toolkit that makes a model behind an OpenAI-compatible endpoint or running in-process answer typed choice, score, and yes/no questions from next-token probabilities in Jev API shape.
Jev Demos & Experiments#community#jevbest#catalog-import
Community
fidecastro/jevifyPython
fidecastro/jevifyPythonFrom 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.