agentbay-monitor-skills

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

• Third-party content exposure detected (high risk: 0.90). The skill's crawl pipeline (scripts/crawl.py and scripts/crawler/* with prompts.yaml) actively fetches public user-generated content from open platforms (小红书/weibo/抖音/知乎 and search engines like Baidu/Bing) into a raw_output_path which the main agent is required to read and interpret per scripts/sentiment/sentiment_instruction.md for sentiment analysis, exposing it to untrusted third‑party content that could carry indirect prompt injection.