ux-brand

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow step "Research references — If WebSearch available" in SKILL.md (and the references/benchmark-framework.md) instructs the agent to generate queries, discover and present 3–5 public-site URLs per reference layer and use those familiarity/ceiling/anti-reference insights to derive design tokens, so it will fetch/read open/public third‑party webpages whose content can materially influence decisions.