transcribe-and-analyze

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

• Third-party content exposure detected (low risk: 0.10). The agent skill downloads media from user-provided URLs and transcribes them using local tools, then optionally reads the resulting transcripts to send to an LLM for analysis. Since media content can originate from arbitrary URLs (like YouTube videos or public media links) whose spoken text is ingested by the agent during analysis, this falls under low risk as the agent actively fetches specific URLs provided.