Skip to content

Jev Browser — tontoko

Playwright wrapper that routes natural-language operations and assertions through Jev while clicks, typing and waits stay deterministic, with verification provenance.

Stars
10
Language
JavaScript
License
APACHE-2.0
Published by
Community
Author
tontoko
Last verified
Oct 2026

This is one of several Jev listings for the same project. The canonical entry is Jev Browser — jkudish.

About this project

Parallel semantic decisions. Deterministic Playwright effects. Explicit verification provenance. Jev Browser combines native Playwright operations with Jev decisions over actual page elements. Use it instead of a Playwright MCP/CLI setup for browser automation, and instead of Stagehand for DOM-grounded act, observe, structured extract, and bounded agent workflows.

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

Highlights

  • Native operations and assertions run with no AI key at all; only the natural-language path touches the model.
  • Batch frontiers share one retained observation across SDK, CLI and MCP, so later native actions need no extra model call while refs hold.
  • assertSemantic re-reads its sources before returning, and evidence that changed or vanished is inconclusive rather than a stale pass.
  • A custom System One endpoint via baseURL or JEV_BASE_URL needs no Jev key and never receives the hosted one.

Watch out

A passing semantic assert is UI evidence, not a database durability guarantee: complete/ui-readback is read that way, unobserved lists fields not shown after saving, and a model's done opinion alone is never treated as completion. A failed or uncertain save is not replayed; an unknown save reconciles read-only first. The package ships as a GitHub Releases tarball rather than a registry install, and it couples tightly to your Playwright version: the peer range is >=1.62.0 <2, an older copy is rejected with a CONFIG error, and Yarn Berry has to add both peers by hand. 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

3GitHub stars
Browser agent where Jev makes every navigation decision and agent-browser performs clicks and typing, stopping when the named goal page is reached or blocked.
Jev Demos & Experiments#community#jevbest#catalog-import
2GitHub stars
Three dependency-free files that put Jev in a browser agent's hot path: code builds the choices from a table of visible controls, Jev picks an operation and target in one forward pass, and the code executes the pick.
Jev Demos & Experiments#community#jevbest#catalog-import
0GitHub stars
A browser MCP server where Jev picks the next page action from locally ranked candidates in roughly 300 ms without LLM tokens, deterministic state diffs verify the step, and the driving agent steps in only on escalation.
Jev Demos & Experiments#community#jevbest#catalog-import
30GitHub stars
Side-by-side game harness where Laya running locally and Jev over API answer the same typed questions for snake, kombat, runner, flappy, tetris, and human-versus-model matches.
Jev Demos & Experiments#community#jevbest#catalog-import
19GitHub stars
MCP server where one browser_goal call hands a URL, a goal, and code-checked assertions to a server-side loop driven by a decision model over live page elements.
Jev Demos & Experiments#community#jevbest#catalog-import
21GitHub stars
Browser top-down driving sim that sends sensor JSON to Jev every 200ms with four typed driving questions in one call, then applies confidence gating and reflex rules to steer.
Jev Demos & Experiments#community#jevbest#catalog-import

From 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
Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai