autoresearch
Warn
Audited by Snyk on Sep 23, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The required workflow instructs the agent to fetch and read third-party social media or podcast content (e.g., X, Reddit) during the research phase without explicit pre-selection of items by the user.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata