customer-discovery
Warn
Audited by Snyk on Feb 17, 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 instructs the agent to search and extract actual quotes and links from public user-generated sources (e.g., Reddit, Facebook groups, Discord servers, public forums via site:reddit.com and similar searches), so the agent will fetch and read untrusted third-party content.
Audit Metadata