The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references a CLI tool installation from belt-sh/cli via npx and links to configuration guidelines on a GitHub repository associated with the inference-sh organization. This behavior is consistent with the skill's purpose of providing access to a specific AI cloud service.
[COMMAND_EXECUTION]: The skill uses the Bash tool to run the belt CLI. The allowed-tools metadata restricts execution to the belt command, which is a common security best practice for limiting tool scope.
[PROMPT_INJECTION]: The prompt parameter accepts user-supplied text which is interpolated into shell commands. While the skill uses single-quoting and JSON formatting in its examples to mitigate command injection risks, this represents a standard surface for indirect prompt injection. The risk is minimized by the platform's command handling.

elevenlabs-music