interview-scheduling
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources (Amplitude user data and Google Calendar events) and interpolates it into Gmail drafts, which creates a potential surface for instructions hidden in that data to influence the agent.
- Ingestion points: Amplitude (session data, feature usage, user profiles) and Google Calendar (existing event details) in
SKILL.md. - Boundary markers: No specific delimiters or safety instructions are used when interpolating external data into the prompt for Gmail draft creation.
- Capability inventory: The skill uses tools to read from Amplitude, read from Google Calendar, and write to Gmail (drafts only).
- Sanitization: No explicit sanitization or validation of the data retrieved from Amplitude or Calendar is performed before use.
Audit Metadata