ai-design-brief

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

• Third-party content exposure detected (medium risk: 0.65). RUNTIME path: the required input fields materialText / materialFile / materialUrl are user-supplied and are composed into the LLM context (x-profile-context: true, contextComposer), and materialUrl can cause the skill to ingest outsider-authored free text from public web pages fetched at runtime.