homeowner-question-content
Warn
Audited by Snyk on Jul 3, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.95). The required runtime workflow explicitly calls UnifAPI to fetch live public web/Reddit/news content (e.g.,
seo/serp→reddit/posts/{id}/commentsandnews/search), which is outsider-authored free text that the agent reads and incorporates into its LLM context.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata