gws-workflow-meeting-prep
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources including calendar event titles, attendee lists, and descriptions.
- Ingestion points: Meeting metadata and descriptions are fetched from Google Calendar via the
gwstool. - Boundary markers: No specific delimiters are used to wrap the ingested content in the output.
- Capability inventory: The skill is strictly read-only and lacks capabilities for command execution, file system writes, or network exfiltration.
- Sanitization: The instructions do not specify sanitization for the ingested data; however, the lack of executable capabilities mitigates the risk of this injection surface.
Audit Metadata