speech-kokoro

Pass

Audited by Gen Agent Trust Hub on Oct 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The 'eval' call identified by static analysis in scripts/kokoro_tts.py is the .eval() method of a PyTorch-based model (KModel). This is a standard operation in machine learning to set a model to evaluation mode and is unrelated to the dangerous Python eval() function that executes strings as code.
  • [EXTERNAL_DOWNLOADS]: The skill downloads model metadata and weights from Hugging Face's official repository (hexgrad/Kokoro-82M). These downloads are part of the intended functionality for local speech synthesis from a well-known AI model hosting service.
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute a local Python script to generate audio. The script's operations are limited to reading input text files, processing audio via the kokoro library, and writing the resulting WAV files to local project directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 18, 2026, 06:23 AM