screenshot

The 'screenshot' skill is designed to capture visual representations of a user's desktop, applications, or specific screen regions at the operating system level. It serves as a fallback mechanism when tool-specific capture options are unavailable or when the user explicitly requests a system-level screenshot. This ensures that the AI can reliably obtain the visual context needed for tasks, inspections, or verification, regardless of the application's native export or capture capabilities.

Key features of this skill include the ability to capture full-screen screenshots, individual windows, or defined pixel regions. Users can specify exact paths for saving captures, rely on OS default locations, or use temporary storage for inspection purposes. On macOS, the skill includes a preflight permission helper to streamline Screen Recording authorization and minimize repeated prompts. The skill supports advanced options such as listing window IDs, capturing active or focused windows, and capturing multiple displays simultaneously. It is implemented with bundled Python and shell scripts, enabling cross-platform functionality for macOS and Linux systems, and integrates seamlessly with workflows requiring automated or repeatable screenshot tasks.

Typical use cases include automated visual inspection by the AI, documenting application states, capturing designs or UI elements from third-party tools, and providing quick visual feedback for troubleshooting or code review. The target users range from developers and designers to AI operators who need consistent, reliable screenshots without manually managing OS-level commands. Its flexibility and automation capabilities make it suitable for scenarios where precision, repeatability, and integration with other tools are important.