What problem does it solve? Automating native desktop applications normally requires stealing the user's cursor and keyboard focus, interrupting their work. This Skill drives the desktop in the background using accessibility-tree element targeting, so GUI automation runs without moving the real cursor, switching virtual desktops, or popping windows to the front. ## Core Features & Use Cases - Element-index targeting: Capture screenshots with numbered overlays on every interactable element (SOM mode) plus an accessibility-tree index, then click by element number instead of unreliable pixel coordinates. - Verify-then-escalate ladder: Every input action returns a structured verdict (confirmed, unverifiable, suspected_noop) with escalation recommendations, so the agent climbs from background element clicks to pixel clicks to foreground delivery only when the driver signals it. - Cross-platform input vocabulary: A single action set (capture, click, drag, scroll, type, key, wait, focus_app) works on macOS, Windows, and Linux with any tool-capable model, powered by cua-driver underneath. - Use Case: Ask the agent to fill out a form in a native desktop app like Figma or Mail while you keep typing in your editor; the agent captures the app, clicks fields by index, types values, and verifies each step without ever taking your focus. ## Quick Start Ask the agent to use computer_use to capture the screen of a specific app and click a named button or link in it.