review-lossless-mapping

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

• Third-party content exposure detected (high risk: 0.85). The skill reads the document under review and the in-scope discussions/decision-log records (which are authored by other participants) as READ-ONLY at runtime, and then emits a review that is based on that outsider-authored free text into the LLM context during steps 4–7.