gcal
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches calendar data and creates events via Google's official API endpoints at googleapis.com.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from the user's Google Calendar.
- Ingestion points: Event summaries, descriptions, and locations are retrieved in scripts/list_events.py and passed to the agent context.
- Boundary markers: No explicit delimiters or instructions are used to separate retrieved data from agent instructions in the output JSON.
- Capability inventory: The agent can execute calendar creation and listing through scripts/create_event.py and scripts/list_events.py.
- Sanitization: The retrieved content is not sanitized or validated before being returned to the agent context.
Audit Metadata