amazon-daily-market-radar

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text enters the LLM context via the review-tagging fallback: apiclaw.py reviews-raw fetches raw Amazon review title/body (outsider-authored) and review-tag-prompt renders them into a prompt string that the caller’s LLM then processes (render_review_map_prompt → Review Text: """{text}""").