| name | apple-spatial-companion-platform |
| description | Implement and validate Apple companion-platform integrations for AirPods/head-tracking workflows, including CoreMotion motion ingest, ear-photo/depth capture pipelines, privacy-retention controls, and platform API boundary decisions. |
Title: Apple Spatial Companion Platform Skill
Document Type: Skill
Author: APC Codex
Created Date: 2026-03-01
Last Modified Date: 2026-03-01
Apple Spatial Companion Platform
Use this skill when work depends on Apple platform APIs in the companion path (Swift/macOS), especially BL-057/BL-058 class slices.
Scope
- CoreMotion integration contracts for
CMHeadphoneMotionManager (availability, auth, device connect/disconnect, fixed-rate behavior).
- Companion capture pipeline for guided left/right/frontal ear-photo inputs and optional depth acquisition when explicitly supported.
- Apple-framework boundary decisions for
CoreMotion, AVFoundation, Vision, and optional ARKit lanes.
- Privacy and retention safeguards for biometric-adjacent image/depth inputs (local processing, no unintended persistence, no network calls).
- Platform capability constraints and non-goals (for example no direct third-party API access to Apple Personalized Spatial Audio profiles).
Required References
references/platform-api-boundaries.md
references/capture-and-privacy-contract.md
references/validation-and-evidence.md
Workflow
- Lock the platform boundary first.
- Confirm what is companion-owned vs plugin-owned for the current slice.
- Keep plugin
processBlock() free of Apple app-level renderer dependencies.
- Confirm API availability and deployment gates.
- Record macOS minimum version assumptions.
- Validate authorization and graceful-degrade paths.
- Define capture contract before implementation.
- Capture sequence, quality gates, and deterministic fallback behavior.
- Output contract for profile-selection inputs (
subject_id, sofa_ref, confidence/fallback reason).
- Implement deterministic processing path.
- Keep embedding/matching behavior reproducible for fixed inputs.
- Keep fallback selection explicit and observable.
- Enforce privacy/retention rules.
- No implicit network calls in capture/match path.
- No raw image persistence beyond explicitly-approved temporary scope.
- Align runbook evidence contracts.
- BL-057: device preset/profile mapping assumptions.
- BL-058: profile acquisition, readiness/sync gating, and axis/frame diagnostics.
- Validate and publish evidence packet.
- Capture runtime status, fallback outcomes, and privacy checks.
- Sync routing/docs when capability claims change.
- Update matrix/runbook references and canonical evidence links.
Cross-Skill Routing
- Pair with
headtracking-companion-runtime for readiness/sync gating and axis/frame diagnostics.
- Pair with
spatial-audio-engineering for renderer/channel/coordinate contract alignment.
- Pair with
hrtf-rendering-validation-lab and perceptual-listening-harness when promotion decisions depend on parity or listening-gate evidence.
- Pair with
skill_docs when backlog/runbook/index wording or governance surfaces must be updated.
Deliverables
- File-level change list with API-boundary rationale.
- Explicit list of privacy/retention guarantees and fallback behavior.
- Validation status:
tested, partially tested, or not tested.