speech-to-text

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

• Third-party content exposure detected (high risk: 0.90). The skill's documentation and workflow (SKILL.md and references/realtime-server-side.md) explicitly show the agent connecting to and streaming/transcribing external HTTPS audio URLs (e.g., the realtime.connect url parameter like "https://npr-ice.streamguys1.com/live.mp3" and the cloud_storage_url option), meaning it ingests untrusted third-party audio that the agent reads and could drive downstream behavior.