speech-to-text

Warn

Audited by Snyk on Mar 6, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.90). This skill explicitly supports streaming and transcribing audio from arbitrary public URLs (see references/realtime-server-side.md "Stream from URL" with url="https://npr-ice.streamguys1.com/live.mp3" and realtime.connect/stream_url examples), so the agent ingests untrusted third-party audio/web content which it must read and interpret as part of its workflow, enabling indirect prompt injection via spoken instructions in those sources.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 6, 2026, 09:25 AM