The Agent Skills Directory

[SAFE]: The skill provides instructions for local text-to-speech synthesis using the kokoro-onnx library, which is a legitimate package for the Kokoro TTS model.
[SAFE]: Dependency installation via pip uses standard, publicly available packages (kokoro-onnx, soundfile).
[SAFE]: File system operations are restricted to writing generated audio to a hardcoded local file path (output.wav), which is standard behavior for a text-to-speech tool and does not allow for arbitrary file writes.
[SAFE]: No prompt injection, obfuscation, persistence mechanisms, or data exfiltration patterns were detected in the instructions or code snippets.

kokoro-tts