The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to invoke ffmpeg and ffprobe for audio processing tasks, such as format conversion (e.g., MP3 to WAV), resampling to 48kHz, and concatenating audio chunks for long-form synthesis. These operations are performed using internally generated temporary filenames and targeted flags.
[EXTERNAL_DOWNLOADS]: Synthesis involves network requests to the official API endpoints of recognized providers including Microsoft Azure, Google Cloud, OpenAI, ByteDance, and Alibaba. These interactions are necessary for the skill's primary function and target well-known, trusted technology services.
[DYNAMIC_EXECUTION]: The backend registry in scripts/backends/__init__.py uses dynamic imports to load the specific adapter for the chosen TTS provider. These modules are local to the skill's distribution and the loading mechanism is used for modular plugin management.
[DATA_EXPOSURE]: The skill maintains a local configuration file (~/.ttsCN.json) and a synthesis cache (~/.ttscn_idem/) to store user preferences and avoid redundant synthesis costs. This follows standard state management practices for CLI-based agent tools.

ttsCN