Eyes and hands on the whole desktop
Screen
When the job is not inside a web page — a QR window, a system dialog, a captcha that watches the mouse — bia can photograph the screen and move like a person.
Screen tools look at the real display, not only Chrome’s HTML. First it takes a picture; then it clicks, types, scrolls, or presses keys. Coordinates come from what it saw, not from guessing.
For suspicious sites it can move the cursor along a curved, human-paced path before clicking. That pairs with Chrome: pages in the browser, everything else on the desktop.
The WhatsApp QR window is a good example — it lives on the display so you can scan it, and screen skills exist for the awkward corners browsers cannot reach alone.
In everyday life
- Takes a screenshot of the whole desktop before acting.
- Clicks, types, and presses keys on whatever is focused.
- Moves the mouse in a human-like path when a site is watching.
- Helps with QR windows, native dialogs, and outside-the-page captchas.
Unlike a normal chatbot
Browser-only agents stop at the edge of the page. bia can reach the rest of the computer when the OS UI is the real obstacle.