parakeet-stt

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized access were detected.
  • [COMMAND_EXECUTION]: The skill uses curl to interact with a transcription API. This behavior is consistent with the stated purpose of a speech-to-text service and targets a local network service.
  • [DATA_EXFILTRATION]: The skill uploads audio files to a user-provided or local network URL (PARAKEET_URL). While this involves data transmission, it is the intended functionality for transcribing audio and does not target sensitive system files or credentials.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes audio data into text. There is an inherent risk that the resulting transcription could contain instructions intended to influence the agent's subsequent behavior if the agent treats the output as trusted content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 08:27 AM
Security Audit — agent-trust-hub — parakeet-stt