competitor-spy

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

• Third-party content exposure detected (high risk: 0.90). The skill's mandatory workflow instructs the agent to perform web_search and web_fetch on competitor URLs, YouTube/TikTok channels, and public forums (see Step 2 Method A "web_fetch [competitor_url]" and Step 2.5 / Step 4 which use web_fetch/web_search and "web_search 'reddit [niche] [product] problems'"), meaning it ingests untrusted, user-generated third-party content that the agent must read and use to drive decisions.