google-calendar
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from the external Google Calendar API, which could include malicious instructions intended to influence the agent.
- Ingestion points: Data enters the context via the
get_eventsmethod inscripts/calendar_client.py. - Boundary markers: Absent. The skill does not use delimiters or instructions to ignore embedded commands in the event data.
- Capability inventory: The skill can perform network operations (Google APIs), read/write local token files, and modify calendar entries (create, update, delete) via
scripts/manage_events.py. - Sanitization: Absent. Event summaries and descriptions are presented to the agent without escaping or validation.
- [COMMAND_EXECUTION]: The skill operates by executing local Python scripts (
setup_auth.py,fetch_events.py,manage_events.py) through shell commands as part of its core functionality.
Audit Metadata