recipe-forward-labeled-emails
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve retrieving the content of external emails and interpolating that content into a forwarding command.
- Ingestion points: The body of Gmail messages retrieved via the
gws gmail users messages getcommand inSKILL.md. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the retrieved email body as untrusted data or to ignore any commands contained within it.
- Capability inventory: The skill uses the
gws gmail +sendcommand, which allows for outbound network communication via email. - Sanitization: The instructions do not specify any sanitization, escaping, or validation of the email body before it is forwarded.
Audit Metadata