create-brand-kit

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

• Third-party content exposure detected (medium risk: 0.30). In Phase 0 the skill performs “WebFetch for copy and positioning” and “WebSearch” on the user-provided company site/starting point, then the extracted “verbatim headline copy” and other mined text are written into brand/BRIEF.md for downstream designer/critic agents to read, making outsider-authored free text ingested at runtime.