The Agent Skills Directory

[SAFE]: The skill implements secure credential management by storing the API key in a configuration file with restricted file permissions (0600) or via environment variables, adhering to security best practices for secret handling.
[SAFE]: Network activity is constrained to the official vendor domain (noiz.ai) for API interaction and downloading generated media, minimizing the risk of unauthorized data transmission.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from lyrics files and external audio sources (files or URLs). Although this creates an indirect prompt injection surface, the risk is low because the processed content is used solely for media synthesis and is not fed back into the agent's instruction pipeline.

text-to-music