recipe-draft-email-from-doc
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a vulnerability surface by processing external, untrusted content from a Google Doc and passing it to a high-privilege action (sending an email).
- Ingestion points: The
gws docs documents getcommand inSKILL.mdfetches content from an external source. - Boundary markers: There are no instructions or delimiters provided to ensure the agent ignores instructions potentially embedded within the document text.
- Capability inventory: The skill uses
gws gmail +sendinSKILL.md, which provides a mechanism for data exfiltration if the document content contains malicious instructions. - Sanitization: No sanitization or validation steps are included for the data retrieved from the document before it is interpolated into the email body.
Audit Metadata