startup-competitors
Warn
Audited by Snyk on Mar 24, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 1.00). The skill explicitly requires live WebSearch and instructs agents in Phase 2 to mine untrusted public, user-generated sources (e.g., G2, Capterra, Product Hunt, App Store reviews, Reddit, Indie Hackers, Hacker News, Quora — see "Wave 2: Customer Sentiment Mining" and related agent templates) and to synthesize those findings into recommendations and battle cards, so third-party content is ingested and can materially influence agent decisions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata