scrollclaw-persona
Warn
Audited by Snyk on Sep 25, 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 agent skill requires mining external product reviews from public sites like Amazon, Trustpilot, and Reddit for persona research, which exposes the LLM to outsider-authored free text.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata