hz-unity-platform-sdk

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate documentation and C# code examples for integrating the Horizon Platform SDK (com.meta.xr.sdk.platform). All URLs point to official developer documentation domains (developers.meta.com, developer.oculus.com) or standard Meta/Oculus API endpoints (graph.oculus.com).
  • [INDIRECT_PROMPT_INJECTION]: The skill describes how to handle DeeplinkMessage and GroupPresenceOptions.DeeplinkMessageOverride, which are ingestion points for external data from other applications or the system. While this represents a theoretical attack surface for indirect prompt injection if the user-supplied string is interpolated into a downstream LLM prompt without sanitization, the skill itself provides standard error handling patterns (msg.IsError) and uses the data for application routing rather than sensitive operations.
  • Ingestion points: ApplicationLifecycle.GetLaunchDetailsRequest() and GroupPresenceJoinIntent.DeeplinkMessage (detected in references/application-lifecycle.md and references/group-presence.md).
  • Boundary markers: Not applicable as this is static documentation.
  • Capability inventory: The skill guides developers to perform application navigation, leaderboard writes, and IAP fulfillment. It does not perform file-system writes or unauthorized network operations itself.
  • Sanitization: The documentation recommends checking IsError and verifying the user identity via Users.GetUserProof() and DeviceApplicationIntegrity.GetIntegrityToken() for high-trust flows.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 06:56 PM
Security Audit — agent-trust-hub — hz-unity-platform-sdk