The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install external components, including the belt-sh/cli via npx and additional skill modules from inference-sh/skills on GitHub. These dependencies originate from the platform provider's repositories.
[COMMAND_EXECUTION]: The skill uses the Bash tool to run the belt CLI. The skill's configuration uses allowed-tools: Bash(belt *) to restrict shell access exclusively to this specific command, which is a security best practice to prevent arbitrary command execution.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes user-supplied text to generate music. This is a primary feature of the skill.
Ingestion points: User-provided prompts are passed via the --input JSON parameter in belt app run commands.
Boundary markers: Data is structured within JSON objects, providing a clear boundary between command logic and user data.
Capability inventory: The skill is capable of executing Bash commands, though restricted to the belt binary.
Sanitization: The skill relies on the external CLI tool to handle and sanitize the input during the inference process.

ai-music-generation