Jev in Practice: Browser Use and Agent Workflows

A short progress update: the team is using Jev with LLMs as an execution harness for world models and calls the results magical, without sharing usage details.

The author argues that Jev costs only $0.00004 per call and that AI removes the upfront capital barrier to entrepreneurship, so the birth lottery no longer caps how far a person can go.

jev-guard uses the Jev engine to risk-score every tool call, blocking destructive actions like rm -rf and flagging prompt injections, with a setup of about five minutes.

Agents read game state from the client's memory, Jev chooses the action plan and GLM writes text, and they negotiate the dungeon and roles without scripts, ending with 6 kills and 0 loot.

Nick Nisi spoke at the AI Engineer conference about using Jev to drive a fruit fly neural network to review code smells, while the author co-ran an agent verification workshop.

The open-source jev-kit uses Jev as a judge rather than a writer, blocking overly expensive sub-agent dispatches and tool calls, completing 9 of 9 browser tasks while cutting Claude spend 233x.

Jev now has vision, and paired with e2e's decision model it can assert on things the accessibility tree cannot see, such as canvas, charts, strikethroughs and drawn text.

The site's "All Web" button launches a Jev-based AI agent that scans five major design inspiration sources, including Pinterest, in real time to find references and top posts by likes.

A video traces Jev's path from a $40M seed round to a $7.5B valuation in 24 days, and OpenAI's fast follow with its Decisions API.

The author shares the agent setup they use every day: Opus 5.5 as the main driver and Jev for fast yes/no judgments at specific points in the workflow.

The author shows how to let Opus plan, Haiku read files, Sonnet fix issues and a verifier agent re-check, while handing deterministic branches like tool choice to Jev, with a prompt to paste in.

The desk screens with free hard rules first, then uses Jev for soft filters like exit risk, momentum and liquidity, as the account grew from $12,155 to $14,853.