topic-agent
Warn
Audited by Snyk on Apr 5, 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). Yes — the SKILL.md explicitly instructs the agent to use WebSearch to fetch and ingest open/public third‑party content (e.g., Reddit posts, Product Hunt entries, blog articles, official blogs) and requires specific article/post URLs as inputs for generating and scoring topics, so untrusted user-generated content can directly influence downstream decisions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata