customer-research
Warn
Audited by Snyk on Aug 26, 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’s “Mode 2: Go Find Research” runtime workflow explicitly directs reading outsider-authored content from online communities/review sites (e.g., Reddit/G2) and extracting verbatim quotes from those sources without requiring the agent to pre-select trusted items first.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata