discovery
Warn
Audited by Snyk on Aug 18, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The skill instructs the agent to actively research outsider-authored free text from sources like Reddit/Hacker News/public forums and specifically mentions “GitHub issues,” “public discussions,” and “YouTube comments,” which are then ingested to extract problem signals.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata