research
Warn
Audited by Snyk on Feb 15, 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's market-research workflow explicitly instructs the agent to fetch and analyze open/public third‑party content (e.g., LinkedIn, Glassdoor, Reddit/industry forums, G2/Capterra, Google News and review sites), meaning the agent will read untrusted user-generated content as part of its routine.
Audit Metadata