compete

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill's Phase 1 research steps (e.g., "Use web search" and Step 1.2) explicitly require fetching competitors' marketing/features pages and user reviews from public sites like G2/Capterra/Reddit (open/public, user-generated sources), which the agent must read and use to drive its analysis and subsequent actions, creating a clear surface for indirect prompt injection.