video-expert-analyzer

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly downloads and parses content from public third‑party sites (e.g., Bilibili/YouTube/Douyin/Xiaohongshu via pipeline_enhanced.py, download_douyin.py, fetch_bilibili_subtitle.py) and then feeds the retrieved transcripts/frames into ai_analyzer.py (Agent/API modes) whose model outputs drive selection, copying of best shots and report generation, so untrusted web/user‑generated content is ingested and can materially influence tool behavior.