Open-source agent-browser launches: browser & desktop automation built for AI agents that verify their own work
evilrabbit_ · x · 2026-11-17
- agent-browser + agent-desktop launch: an open-source browser/desktop automation toolkit built for AI agents, designed to plug into your existing coding agent.
- Core capability: lets agents click, type, and complete tasks in the browser; agent-desktop extends the work into native apps (demo shows Gemini, OpenAI, and Claude agents operating a desktop simultaneously).
- Verify-then-continue workflow: in the official demo, an agent spots a typo on a production site, screenshots it, opens the fix's Preview deployment, screenshots that too, pastes both Before/After images into the PR description, and closes out — all without human intervention.
- Other features: compact page-control snapshots to cut token usage, region-level window mapping, screenshots and up to 60fps recordings, Core Web Vitals performance tests, and network mocking.
More from coding & agent
- Ripple-65B hackathon: model builds its own IDE with <80ms autocomplete, 300 tok/s — Azaliamirh · 2026-11-17
- LM Studio launches Agent Remote Control for running agents on any device — MatthewBerman · 2026-11-17
- Qwen 3.8 27B at 300-500 tok/s on a 5090 laptop, tuned for parallel agents — tebjan · 2026-11-17
- Ripple-65B Hackathon: Self-Quantization Shrinks Model from 60GB to 30GB with Little Quality Loss — Azaliamirh · 2026-11-17
- Treating agent orchestration like a factory: the 7 jobs of an orchestration layer — blaizedsouza · 2026-11-17
- UCLA and AWS paper: AI agents find math shortcuts but thinking time often cancels the gains — rohanpaul_ai · 2026-11-17