The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the belt CLI tool to perform its primary functions, including authentication (belt login) and running AI models (e.g., belt app run infsh/kokoro-tts). These operations are within the scope of the skill's intended purpose for voice synthesis.
[EXTERNAL_DOWNLOADS]: The skill directs users to install external software components, specifically the belt-sh/cli package via npx and configuration files from the inference-sh GitHub organization. These resources originate from the official infrastructure of the service being integrated.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a conduit for user-supplied text that is processed by remote AI models.
Ingestion points: User-provided strings in the prompt and text fields of the JSON payloads passed to the belt CLI (SKILL.md).
Boundary markers: The use of structured JSON for tool inputs helps separate user-provided content from the command structure.
Capability inventory: The skill utilizes the Bash tool to execute CLI commands and manage local audio/video files resulting from the synthesis.
Sanitization: Content is passed as a data payload to the inference.sh API, which manages its own safety filtering for model inputs.

ai-voice-cloning