whisper
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (SAFE): The skill transcribes audio input which could contain embedded instructions (Indirect Prompt Injection). This is an inherent risk of transcription capabilities and is not a flaw in the skill's implementation.
- EXTERNAL_DOWNLOADS (SAFE): The library downloads model weights from OpenAI's official infrastructure. Per trust rules, downloads from trusted sources for primary functionality are considered safe.
Audit Metadata