ad-copywriting
Warn
Audited by Snyk on Jul 27, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.75). This skill’s required workflow explicitly tells the operating user to gather voice-of-customer data from outsider-authored sources (e.g., G2/Capterra/TrustRadius reviews, sales call transcripts, support tickets, Reddit/LinkedIn posts), and that extracted free-form text is then used by the LLM to write ad copy, creating an indirect prompt-injection path through those externally authored quotes.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata