Post · Agents & browsers
Computer use without screenshots
On-device segmentation and OCR turn the screen into text for Jev.
Open x.com ↗How builders describe it
Computer use with Jev and local vision model! Its so goddamn fast
The decision Jev makes
Pick the next click, field or navigation from the legal actions on screen.
Where it fits
Browser agents need to choose an action on every step. With Jev each step is a choice from the actions actually available on the page, which keeps agents fast and on the rails. All 193 agents & browsers projects →
Related projects
Agents & browsersjev-kitEverything you need to run TypeSafe's Jev with Claude Code: a tool-call guard, tier guard, file search, browser agent, review, belay, compaction and…GitHub · ★ 25 · jonathanavis96
Agents & browsersjev-chat-windowsWindows 微信回复助手（非官方修改版）：本地离线 OCR 读屏 + 模型起草候选回复 + 一键填入微信输入框，可选自动发送。不 hook、不注入、不读微信数据库。GitHub · ★ 5 · caizili999
Agents & browsersCUA-JEVOpen reference framework for Jev-powered computer use: task adapters propose legal actions; TypeSafe Jev chooses a concrete action and channel; code…GitHub · ★ 3 · ZJU-REAL
Agents & browserswevLocal, open-weights decision model for browser agents: a drop-in for System One / Jev requests.GitHub · ★ 3 · alanhuangyoo
Agents & browsersjev-browser-skill (ChenYCL)Browser and computer-use CLI/MCP for coding agents: TypeSafe Jev supplies calibrated action judgments while code owns the control loop (ego lite /…GitHub · ★ 2 · ChenYCL
Agents & browsersjevbriefPlaywright briefing layer for TypeSafe Jev: filters page elements with drop reasons, asks one Choice for the next click, and writes a local trace…GitHub · ★ 2 · Parthkomalwad