analyze

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md step 4–5 (“Survey the input” and “Discover embedded sources”) requires reading the user-supplied source artifact(s) (including referenced embedded markdown/sibling .md files) to draft tour-spec.json, so outsider-authored free text is ingested via the runtime’s artifact-path workflow.