What problem does it solve? Automating macOS desktop tasks normally requires manual clicking, typing, and screen inspection. This Skill gives an AI agent full control over the macOS UI through the Peekaboo CLI, enabling screen capture, element targeting, input simulation, and app/window management from the command line. ## Core Features & Use Cases - Screen Capture & Vision: Take screenshots of screens, windows, or menu bar regions, generate annotated UI maps with see, and optionally analyze captures with AI prompts. - UI Interaction: Click, type, drag, swipe, scroll, and press hotkeys by targeting elements via snapshot IDs, coordinates, or app/window identifiers. - System Management: Launch, quit, hide, and switch apps; manage windows, Spaces, Dock items, menus, menu bar extras, dialogs, and the clipboard. - Use Case: Automate a login flow by running peekaboo see --annotate to map the UI, then peekaboo click --on B3 and peekaboo type "user@example.com" to fill in credentials without manual input. ## Quick Start Ask the agent to run peekaboo permissions, then use peekaboo see with the annotate flag to capture the current screen and identify clickable elements before automating any clicks or typing.