amazon-competitor-intelligence-monitor

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

• Third-party content exposure detected (high risk: 0.85). The runtime path that feeds free text into the agent’s LLM context is the review fallback chain: zoodata.py review-tag-prompt renders prompts containing raw review body text returned by the realtime/reviews API, and the calling LLM then ingests that review text to produce tags (outsider-authored Amazon reviews).