competitive-research

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly requires web research that fetches and ingests publicly available, untrusted third-party content (e.g., web search, competitor product pages, review platforms like G2/Capterra, Product Hunt, social media, and community forums such as Reddit/Hacker News) as part of its required workflow, so that content can materially influence the agent's analysis and actions.