SuperAPI AI Directory

arkit-camera-access-providers

Agents & models

View repository

Updated 10-16 08:49

About this skill

ARKit Camera Access Providers

Quick Start

  1. Confirm the task needs ARKit camera data, not a rendered RealityKit or SwiftUI snapshot.
  2. Choose full camera frame access or region-scoped camera capture.
  3. Check support, privacy text, entitlement, and enterprise constraints before debugging update delivery.
  4. Keep frame handling off the SwiftUI body path and avoid retaining stale frame buffers beyond the API contract.

Load References When

Reference When to Use
session-basics.md When setting up ARKitSession, authorization, events, or teardown.
camera-frame-provider.md When using CameraFrameProvider and cameraFrameUpdates(for:).
camera-region-provider.md When using CameraRegionProvider and CameraRegionAnchor.

Workflow

  1. Verify the camera-access requirement and target OS/device support.
  2. Confirm Info.plist text, entitlement/profile, and enterprise requirements.
  3. Select camera format or region definition deliberately.
  4. Request provider authorizations, run the provider, and process updates in a dedicated model or pipeline object.

Guardrails

  • Do not assume all camera providers are generally available; treat missing entitlement/profile support as a first-class blocker.
  • Do not widen camera access when a narrower region or non-camera API satisfies the feature.
  • Do not route camera provider build, signing, or provisioning failures here; use $signing-entitlements or $build-run-debug.

Output Expectations

Provide:

  • which camera provider was chosen and why
  • support, privacy, entitlement, and enterprise checks
  • which camera references were used
  • the frame or region processing model