topic-collector
Warn
Audited by Snyk on Feb 16, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 1.00). The skill explicitly uses WebSearch to fetch and summarize content from public, user-generated sources—e.g., Twitter/X, Reddit, Hacker News, Product Hunt and arbitrary blog URLs—and requires extracting and interpreting those external posts/links as part of its workflow, exposing the agent to untrusted third-party content.
Audit Metadata