create-value-prop-ad

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

• Third-party content exposure detected (high risk: 0.75). Yes—Phase 0 “Autonomous brand intake” scrapes free-form text (homepage H1/tagline, meta description, PDP bullets, footer trust strip, stick-pack label copy) from the runtime-supplied brand_url, and that scraped prose is then used to populate brief.md/shot-list.yml and ultimately rendered into beat HTML (LLM context risk via any LLM step that consumes brief.md/scraped props).