speech-to-text
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive documentation for integrating the ElevenLabs Speech-to-Text API.
- [EXTERNAL_DOWNLOADS]: The skill recommends installing official, well-known packages including
@elevenlabs/elevenlabs-js,@elevenlabs/client, and@elevenlabs/reactfor Node.js, andelevenlabsfor Python. These are standard dependencies for the service. - [CREDENTIALS_UNSAFE]: The skill correctly instructs users to manage sensitive credentials like
ELEVENLABS_API_KEYusing environment variables or backend token generation, following security best practices to avoid exposing keys in client-side code. - [DATA_EXFILTRATION]: Network operations are limited to communicating with official ElevenLabs API endpoints (
api.elevenlabs.io) or retrieving audio from provided example URLs. There are no patterns suggesting unauthorized data exfiltration. - [INDIRECT_PROMPT_INJECTION]: While the skill processes external audio data for transcription, it does not include instructions to execute or further process the resulting text in a way that would trigger a vulnerability. The severity for this inherent capability is low.
Audit Metadata