afrexai-competitive-intel

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md workflow explicitly instructs the agent to fetch and interpret open web and user-generated sources (e.g., Google, G2/Capterra/TrustRadius reviews, Product Hunt, Crunchbase, LinkedIn, Reddit/Hacker News, competitors' pricing pages and the Wayback Machine) and then use that intel to drive product, pricing, and battlecard decisions, which exposes the agent to untrusted third-party content that can influence actions.