The Agent Skills Directory

[SAFE]: The skill implements strong security defaults for credential management, explicitly instructing users to utilize environment variables for all API keys and tokens. It includes warnings against hardcoding secrets in multiple reference files.
[EXTERNAL_DOWNLOADS]: The documentation provides instructions for setting up local audio models and integrating with cloud services. These references target well-known and trusted organizations such as Meta (AudioCraft), Suno AI (Bark), OpenAI, Google Cloud, and ElevenLabs. Installation commands use official package registries (PyPI, npm) or official GitHub repositories.
[COMMAND_EXECUTION]: The skill generates shell scripts and Python/JavaScript code to automate audio processing tasks using 'ffmpeg'. These operations (LUFS normalization, format conversion, and silence trimming) are standard industry practices for game development and are executed within a clearly defined scope.
[INDIRECT_PROMPT_INJECTION]: As an agent that processes natural language prompts to generate audio parameters, it technically possesses an ingestion surface. However, the risk is mitigated by the fact that the output is audio-specific code and the instructions include boundaries that prevent the execution of arbitrary commands derived from prompt data.

tone