kuaishou-comment

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text from public Kuaishou comments is fetched at runtime via scripts/kuaishou_comment_search.py (Redfox API) and then ingested into the agent’s LLM context when the workflow renders “评论表格（展示当前页全部评论）” and uses those comment texts for “AI 评论总结”.