light-review-rebuttal

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md’s required runtime path includes capturing public OpenReview forum text via scripts/fetch_openreview.py (raw reviewer/comment content/comment becomes raw_text), which is outsider-authored free text that is then embedded into the workflow’s generated artifacts and thus can be included in the agent/LLM context downstream.