transcribe

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the official openai Python package, which is a standard dependency for accessing OpenAI services.
  • [COMMAND_EXECUTION]: The skill executes a bundled Python script (transcribe_diarize.py) to process audio files and interact with the OpenAI API.
  • [DATA_EXFILTRATION]: The skill reads local audio files and optional speaker reference files to transmit them to OpenAI's transcription service. This data transfer is the intended core functionality and targets a well-known, established service provider.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 03:34 AM
Security Audit — agent-trust-hub — transcribe