01Evaluates and analyses what’s rendered on screen to decide the next action
02Uses native accessibility trees and screenshot-based coordinates for interactions.
0346 GitHub stars
04LLM-friendly, no computer vision model required in Accessibility (Snapshot).
05Offers a comprehensive suite of mobile commands for interacting with devices and applications.
06Extract structured data from anything visible on screen.