paw-mkt-guerrilla

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly includes a "live URL" starting mode and Response Protocol step to route to a "live URL audit (existing funnel or campaign review)" — indicating the agent will fetch and read arbitrary external webpages/social feeds (and the references also instruct monitoring public platforms like Reddit, X, Product Hunt), so it will ingest untrusted third-party content that can influence decisions.