Jev in Practice: Browser Use and Agent Workflows

Frames from the glasses reach a Mac over a WhatsApp video call, where Jev cheaply pre-checks readability and board changes before Opus 5.5 reads the position and Stockfish picks a move spoken back through a virtual mic.

The author built a small side project with Jev to connect with founders, engineers, AI people, browser users, and design enthusiasts.

poor-things.com runs 67 prompts per steering config, scores outputs with Typesafe Jev and Fireworks Rune, then has GLM 5.3 Flash name the personalities.

A developer says he wires Jev into every tool he builds for faster decisions, and finds its new image input strong for visual judgment and highlight filtering.

In a 12-agent Halloween costume experiment, Jev is the one that builds and ships while Opus directs, reportedly earning $7,237 in one night.

Yuvelir_22 describes wiring six AI tools into a night-shift system where Jev builds and ships the product, and by morning it had sign-ups and paying users.

An open-source workflow where Jev cheaply pre-checks frames to skip unchanged boards, Opus 5.5 reads the position only when needed, and Stockfish suggests moves spoken into the wearer's ear.

A small local demo routes connections between two papers into look-closer, send-for-review or leave-gap outcomes, with Jev handling the sorting and Claude investigating promising leads.

TypeSafe claimed all customer outputs leaked and linked to 101 probability values from 0.00 to 1.00, a prank built on Jev decision outputs.

A trading dashboard rendered as a virtual trading floor with desk characters, live charts and positions, showing $37,085 across five desks.

The author maps protein-folding timing onto agents: before each step, hand Jev the current state and get back a structured KEEP GOING / GET CONTEXT / REVIEW choice that code routes on.

Six AI tools including Opus, Grok, and Jev are chained into a team that finds a need, builds a product, and promotes it unattended for 12 hours, claiming 47 paying customers.