alpha-insights
Warn
Audited by Snyk on Jul 1, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.65). Outsider free text can enter the LLM context via Track E “Social Media” / Track G “User Voice” during Stage 4 research execution, where runtime-fetched posts/comments or user-provided messages from non-operating parties are ingested as readable evidence text.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata