About
The vision-ref skill is an essential technical companion for xOS developers implementing computer vision capabilities. It provides instant access to detailed API signatures, landmark mapping, and implementation patterns for Apple's Vision and VisionKit frameworks. From subject lifting and instance segmentation to complex 2D/3D human body and hand pose detection, this skill ensures developers follow Apple's best practices for coordinate mapping and performance optimization. It is particularly useful for building gesture-driven interfaces, augmented reality effects, and sophisticated image analysis tools across iOS, macOS, and visionOS.