transcribe

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill depends on the openai Python library, which is a well-known and trusted package from a recognized organization.
  • [SAFE]: The skill implements secure credential management by instructing users to export their OPENAI_API_KEY as an environment variable rather than prompting for it to be pasted into the chat interface.
  • [SAFE]: The bundled Python script scripts/transcribe_diarize.py performs standard audio file reading and API communication. It includes appropriate file validation and size checks, and it does not contain any patterns indicative of command injection, data exfiltration to unauthorized domains, or persistence mechanisms.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 06:56 AM
Security Audit — agent-trust-hub — transcribe