recipe-audit-external-sharing
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes untrusted data from an external environment.
- Ingestion points: The skill reads file metadata and permission lists from Google Drive via the
gws drive files listandgws drive permissions listcommands. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious content embedded within file names or metadata.
- Capability inventory: The skill includes the capability to delete permissions via the
gws drive permissions deletecommand, which represents a state-changing operation on cloud resources. - Sanitization: No sanitization or validation of the retrieved metadata is performed before the agent uses the information to construct subsequent commands.
Audit Metadata