deepgram-data-handling

Warn

Audited by Snyk on Feb 20, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.70). The skill explicitly ingests and processes untrusted user-provided audio (see the SecureAudioUpload upload(audioBuffer, ...) example and the transcribeWithRedaction call to client.listen.prerecorded.transcribeUrl({ url: audioUrl }) in SKILL.md), causing the agent to read/interpret third-party/user-generated content that could influence downstream handling or actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 20, 2026, 05:23 AM