deepgram-java-audio-intelligence

Warn

Audited by Snyk on Jul 16, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.75). The required workflow is a Deepgram Listen V1 transcription request (URL/byte upload) whose returned transcript/analysis is external service output (outsider-authored free text) that would be ingested into the agent/LLM context via the SDK response handling path.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 16, 2026, 09:07 PM
Issues
1
Security Audit — snyk — deepgram-java-audio-intelligence