alarmkit
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions and reference implementation patterns ingest untrusted user data via metadata structures which are then interpolated into system UI components.
- Ingestion points: Metadata fields such as
label,recipeName, andstepDescriptioninreferences/alarmkit-patterns.md. - Boundary markers: None identified in the prompt templates or implementation patterns.
- Capability inventory: The skill utilizes
AlarmManagerto schedule system-level alerts and Live Activities that can bypass Silent and Focus modes. - Sanitization: No evidence of data validation or sanitization before interpolation into UI components.
Audit Metadata