APIClaw — Amazon Commerce Data, 11 Endpoints

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches live, user-generated Amazon review text via the realtime/reviews endpoint and instructs the calling LLM to run Map/Reduce on those raw reviews (see "Local Review Toolkit" in SKILL.md and render_review_map_prompt / fetch_realtime_reviews_all in scripts/apiclaw.py), so untrusted third‑party content is ingested and directly influences downstream analysis and agent decisions.