arkit-face-body-tracking-workflowlisted

Guide ARKit TrueDepth face tracking, face anchors, geometry, topology, transforms, eye transforms, blend shapes, face-driven animation, world-plus-face tracking, body anchors, skeletons, joint transforms, scale estimation, privacy, and diagnostics. Use when ARKit face geometry or 3D body tracking is primary rather than Face ID authentication or Vision frame analysis.
gaelic-ghost/socket · ★ 6 · AI & Automation · score 69

Install: claude install-skill gaelic-ghost/socket

# ARKit Face and Body Tracking Workflow ## Purpose Guide ARKit face and body tracking while keeping TrueDepth geometry, body skeletons, Vision image analysis, Local Authentication, camera capture, rendering, and sensitive-data handling explicitly separate. ## When To Use - Use for `ARFaceTrackingConfiguration`, `ARFaceAnchor`, face geometry, blend shapes, eye transforms, `ARBodyTrackingConfiguration`, `ARBodyAnchor`, `ARSkeleton3D`, joint transforms, and scale estimation. - Recommend Local Authentication for Face ID or Touch ID authentication; ARKit never exposes enrolled biometric templates. - Recommend Vision for face rectangles/landmarks or 2D body/hand/animal pose analysis in ordinary images and video frames. ## Single-Path Workflow 1. Classify the request: - face position/orientation or topology - blend shapes or face-driven animation - eye transforms or gaze-adjacent effects - world-plus-face tracking - body anchor, skeleton, joint pose, or scale - rendering, privacy, performance, or correctness repair 2. Apply the Apple docs gate: - read current ARKit face/body and Local Authentication documentation - state the documented behavior relied on - apply `../../shared/references/apple-spatial-data-privacy-contract.md` - check configuration support, supported video formats, maximum tracked faces, world-tracking support, automatic scale support, and physical device capability 3. Choose one tracking configuration: - use `ARFaceTrackingConf