recipe-draft-email-from-doc
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill demonstrates a workflow that ingests untrusted data from an external source and passes it to a communication tool, creating a potential injection surface.
- Ingestion points: Data is retrieved from an external Google Document via the
gws docs documents getcommand inSKILL.md. - Boundary markers: The skill lacks delimiters or instructions to the agent to ignore any malicious instructions that might be embedded within the retrieved document text.
- Capability inventory: The skill utilizes the
gws gmail +sendcommand, which allows for network-based data transmission. - Sanitization: No sanitization, escaping, or validation logic is defined to inspect the document content before it is placed into the email body.
Audit Metadata