Adversarial browser testing using Jev and Runlayer

X postRafal WilinskiDemo

Rafal Wilinski showcases a testing suite that employs parallel AI agents and Chromium to break releases. The setup relies on TypeSafeAI's Jev model and Runlayer agents.

Key points

  1. Wilinski uses TypeSafeAI's Jev model for browser-based adversarial testing across releases.
  2. The setup combines Runlayer agents with Chromium to run tests in a massively parallel fashion.
  3. Instead of raw visual data, the agents process a simplified, transformed HTML tree.
  4. Wilinski claims the testing suite operates at negligible cost ('costs pennies').

Why it matters

Browser-based adversarial testing has traditionally been costly, fragile, and difficult to scale across every build. Running cheap, agentic testing pipelines allows development teams to continuously stress-test web applications before deployment.

Analysis

By feeding transformed DOM trees rather than raw visual screenshots into the agents, Wilinski keeps token usage and latency low. The combination of Runlayer's execution environment and TypeSafeAI's Jev model illustrates how targeted LLM tooling can make continuous adversarial end-to-end testing economically viable.

Original post

Rafal Wilinski

@rafalwilinski · 18 September 2026

thanks to @typesafeai's Jev we now have massively parallel browser-based adversarial testing suite that tries to break each release. and it costs pennies. t.co/WoIlqGjjNB

4.4k likes82 replies3.4k bookmarks238.5k views

View on X
More posts in the thread (3)
  1. @jedfrankowski @typesafeai @runlayer agents + chromium
  2. Since everyone is asking about setup - it’s just @runlayer agents + chromium magic. Dogfooding at its finest
  3. @ashleyrudland @typesafeai simplified/transformed html tree
Selected replies (5)
  • @tommyjmarshall @rafalwilinski @typesafeai What definition of “break” are you using? How can it tell between something visually regressing a little?
  • @air_codex @rafalwilinski @typesafeai holy moly the jev hype is real
  • @ashleyrudland @rafalwilinski @typesafeai You can’t give dev images so how does this work?
  • @Richelle_Ji @rafalwilinski @typesafeai Open-weights jev exists now, with vision: t.co/yMmnOYhXVO t.co/j4HfXEiU8b
  • @HaseebMir91 @rafalwilinski @typesafeai Why it was not possible before Jev model? i am curious

Mentioned links

Links and tools from this entry

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Repository:nekuda-ai/WindTunnel

    WindTunnel: Benchmarking Browser Agent Interfaces and WebMCP

    WindTunnel is an open-source benchmark suite designed to compare how browser agents interact with websites. It evaluates WebMCP (direct site-exposed actions) against coordinate-based computer use via screenshots and page structure navigation using DOM and accessibility trees.

    34starsHTML

    AI & LLMs· Research

  • Video

    Video:AI Engineer

    Web Automation with AI Agents via the Chrome DevTools Protocol

    Corey Gallon demonstrates how AI agents can bypass modern bot detection systems like Cloudflare Turnstile and reCAPTCHA using the Chrome DevTools Protocol (CDP). By combining deterministic execution with targeted vision calls and human-like mouse dynamics, agents interact with browsers indistinguishably from human users.

    AI & LLMs· Talk

  • Video

    Video:AI LABS

    Graph Engineering for AI Agents: Parallel Workflows and Verification

    Graph engineering supersedes linear loop engineering by splitting complex tasks across multiple parallel sub-agents. AI Labs outlines the node-and-edge architecture, the risk of cascading failures, and concrete verification patterns implemented with Claude Code.

    AI & LLMs· Tutorial

  • Article:Google Chrome Developers Team

    Lighthouse Introduces Experimental Audits for Agentic Browsing

    Google has documented an experimental Lighthouse category designed to evaluate how well websites interact with autonomous AI agents, using pass/fail checks rather than a 0-100 score.

    AI & LLMs· Resource

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.