idea-discovery
Warn
Audited by Snyk on Apr 7, 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). Mode 2 "Outbound — Competitive Research" explicitly instructs the agent to research competitor homepages, pricing, changelogs and public sites like G2/Reddit, so the agent will fetch and interpret untrusted public third-party content that can materially influence idea generation and subsequent actions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata