ad-creative-intelligence

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs running meta-ad-scraper and google-ad-scraper to scrape Meta and Google ad libraries and landing page URLs (public third‑party ads and webpages), which the agent ingests and analyzes to drive clustering and creative recommendations, so untrusted third‑party content can materially influence actions.