transcribe

Pass

Audited by Gen Agent Trust Hub on May 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses the official and well-known openai library for all transcription and diarization functionality.
  • [SAFE]: Sensitive information like the OpenAI API key is handled through standard environment variables, and the instructions explicitly advise against insecurely sharing keys in chat.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted audio data which could contain malicious spoken instructions.
  • Ingestion points: Audio files are read and processed by the scripts/transcribe_diarize.py script and passed to the OpenAI API.
  • Boundary markers: No specific delimiters or boundary instructions are used to isolate the resulting transcript within the agent's prompt context.
  • Capability inventory: The bundled Python script has permissions to read local files, write transcripts to disk, and make outbound network requests to the OpenAI API.
  • Sanitization: The skill does not perform sanitization or validation on the content of the audio files or the generated text output.
Audit Metadata
Risk Level
SAFE
Analyzed
May 17, 2026, 03:02 PM
Security Audit — agent-trust-hub — transcribe