Bookmarks
Warn
Audited by Snyk on Apr 20, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs connecting and importing saves from public, user-generated platforms (e.g., X, YouTube, Reddit, Pinterest, Instagram, TikTok) — see the "Setup" and "Sources" sections — and the agent ingests and acts on that content (auto-tagging, summaries, alerts), so untrusted third-party content could influence behavior.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata