activitykit

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documents the handling of external push notification payloads that update the agent's display state.
  • Ingestion points: aps.content-state JSON payloads received via APNs, as described in references/lifecycle-updates-and-push.md.
  • Boundary markers: Strict enforcement of Swift Codable protocols ensures incoming data matches the pre-defined ContentState schema.
  • Capability inventory: The skill is limited to updating visual elements on the Lock Screen and Dynamic Island; it does not possess capabilities for file modification, arbitrary command execution, or sensitive data access.
  • Sanitization: The documentation explicitly advises developers not to include secrets in attributes or state and relies on the framework's internal sandboxing for display.
  • [EXTERNAL_DOWNLOADS]: The skill contains references to external documentation and testing tools.
  • Provides links to a documentation viewer at sosumi.ai.
  • Includes a curl example for testing push updates against Apple's official APNs gateway (api.push.apple.com).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 11:13 AM
Security Audit — agent-trust-hub — activitykit