customer-email-draft-threads
Warn
Audited by Snyk on Jul 28, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.95). Yes—runtime Gmail thread bodies are treated as untrusted input and are explicitly “present[ed]” to the model via the required chronological
Conversation threadreadback before drafting/handoff, meaning outsider-authored email/free-text from customers is ingested into LLM context.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata