Skip to content

Jev UI

React library on npm where Jev judgments choose presentation: `Branch` picks a subtree, `Rank` orders or truncates a candidate set, `Gate` shows an affordance, and `useScore` returns a number you can use as an ordinary prop.

Stars
5
Language
TypeScript
Published by
Community
Author
etweisberg
Last verified
Oct 2026

About this project

React components that decide which component to render, how to order a list, and whether to show an affordance — from calibrated judgments returned by TypeSafe's Jev. The model returns a typed answer with probabilities and a confidence. Your code keeps the thresholds, the fallbacks, and every action. Create one in the TypeSafe console. It looks like apikey-.... Create .env in the root of your Next.js app: Add it to .gitignore before you paste the key in:

Quoted from the project's README. Read the full README.

Highlights

  • Thresholds, fallbacks and every action stay in your code; the model returns a typed answer with probabilities, and `Gate` only adds content.
  • Every judgment in one render pass is one request, because state is ingested once for all of them, and each component reports its cost through `onResolved`.
  • Numeric telemetry is projected to words before the model sees it: `41200 ms` becomes `time_spent: "long"`, since Jev underperforms on numeric representations.
  • Four environment variables cover the common case, including `JEV_TRANSPORT` with live, record, replay, mock and auto, and `JEV_FIXTURES_DIR`.

Watch out

This decides presentation only; nothing in the library submits a form, spends money or deletes a row, and the `when` text on each `Option` becomes the model's criteria, so a bare key like `"chart"` gives it nothing to reason about. Hiding content behind a judgment is called out as an accessibility and SEO problem, which is why `Gate` only adds. The demos are described as live, but running the site yourself with your own key is what makes each demo call the real API. 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.

More Jev resources like this

69GitHub stars
Provider-neutral TypeScript runtime for System One decisions: write choice, noul, and score code once, then run it on TypeSafe Jev, local Reflex, any compatible /v1/systemone endpoint, or a custom provider.
Jev Integrations & Adapters#community#jevbest#catalog-import
15GitHub stars
Uncertainty-aware semantic control flow and routable procedures for Effect, with type-safe patterns and policies over Effect Decision and DecisionModel where uncertain answers stay explicit branches.
Jev Integrations & Adapters#community#jevbest#catalog-import
12GitHub stars
Command-line checker that asks Jev whether each selected source file violates a plain-English convention, returning file-level probabilities for an agent or CI step to act on.
Jev Integrations & Adapters#community#jevbest#catalog-import
8GitHub stars
Shadcn registry search block that shows lexical hits instantly then re-ranks the top candidates with Jev intent scoring in one request, with keyword fallback on failure.
Jev Integrations & Adapters#community#jevbest#catalog-import
2GitHub stars
A Next.js workbench that converts OpenAI-compatible Chat Completions requests into Jev state and typed questions, compares chat output against structured judgments, and publishes saved question packs as proxy routes. The interface is mainly in Chinese.
Jev Integrations & Adapters#community#jevbest#catalog-import
n8n community node for TypeSafe's System One API, exposing yes or no, choice and score questions as workflow nodes, with an Evaluate Questions operation that batches several over one state in a single request.
Jev Integrations & Adapters#community#jevbest#catalog-import
Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai