apply-comments

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

• Third-party content exposure detected (high risk: 0.75). The required workflow reads slides/<slideId>/index.tsx and parses /* @slide-comment ... text="...base64url(JSON)..." */ markers (outsider-authored slide comments) into free-text note/hint that is then used to drive LLM reasoning/editing, creating an indirect prompt-injection path via the slide file’s embedded comment text.