What problem does it solve? Building camera and audio features on Meta smart glasses is error-prone: SDK symbols change between releases, phone microphones get mistaken for glasses microphones, and mock-device tests get passed off as physical-device proof. This Skill enforces exact SDK-version verification, explicit stream ownership, consent handling, and separated evidence tiers so wearable media features are built against the real API contract. ## Core Features & Use Cases - Camera workflow guidance: Gate on registration, session readiness, and permissions; configure resolution, frame rate, and format from supported values; own the stream through Camera.stop() with bounded buffering and cancellation on lifecycle events. - Audio route discipline: Distinguish phone microphone, HFP glasses microphone, and A2DP playback; configure AVAudioSession for the documented route with consent, retention, and fallback policies defined up front. - Evidence separation: Require mock, simulator, connected-device, and physical-device test evidence so a phone-camera or MockDevice pass never substitutes for glasses hardware proof. - Use Case: When adding a glasses camera preview to an iOS app on DAT 0.9.0, use this Skill to verify DeviceSession.addCamera(config:) symbols, scaffold from the camera starter adapter, and produce a permission matrix plus privacy data-flow table before release. ## Quick Start Ask the agent to design a glasses camera capture flow for your Meta DAT iOS app and produce the capability matrix, ownership diagram, and device evidence plan.
