diagram-design-editorial

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

• Third-party content exposure detected (high risk: 0.80). The skill's Brand Onboarding workflow explicitly instructs Claude to "fetches homepage" and "extracts dominant palette + font stack" from arbitrary sites (see "Brand Onboarding" in SKILL.md), so it ingests content from open/public third‑party webpages which directly influences diagram generation and token/behavior decisions.