recipe-draft-email-from-doc

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from a Google Document and feeds it into an outbound channel (Gmail) without safety boundaries or sanitization.
  • Ingestion points: SKILL.md (Step 1) fetches content from a Google Doc using gws docs documents get.
  • Boundary markers: Absent. There are no boundary delimiters or instructions telling the model to ignore potential malicious prompts inside the fetched content.
  • Capability inventory: SKILL.md (Step 3) possesses the capability to send emails via gws gmail +send.
  • Sanitization: Absent. The text extracted from the document is directly embedded into the email body without validation or filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 11:16 PM
Security Audit — agent-trust-hub — recipe-draft-email-from-doc