The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads text-to-speech models (~335MB) from a public GitHub repository (github.com/nazdridoy/kokoro-tts) during its first execution.
[COMMAND_EXECUTION]: The skill uses subprocess.run and subprocess.Popen in scripts/read_aloud.py to install the kokoro-tts tool via uv and to execute the speech synthesis engine.
[COMMAND_EXECUTION]: The SKILL.md file contains dynamic context injection markers (!command) used to perform benign environment checks, such as verifying if uv or the voice models are installed, to provide status information to the agent.

read-aloud