xiaoyuzhou-podcast-transcriber

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to fetch the Xiaoyuzhou episode page HTML and extract/download the episode audio from public Xiaoyuzhou pages (第一阶段：提取音频 URL / 下载音频), then transcribe and summarize that user-generated podcast content—so untrusted third‑party content is ingested and directly shapes the agent's outputs.