The Agent Skills Directory

[SAFE]: The skill interacts with the vendor's official infrastructure (muapi.ai) to perform its primary function. It uses standard practices for configuration, such as sourcing environment variables for API authentication, and does not exhibit patterns of data exfiltration to unknown third parties.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it processes external video content and metadata returned by an API.
Ingestion points: The --source video URL or local file provided to the script, which is then processed by the muapi.ai service.
Boundary markers: The skill enforces a strict JSON schema for the output (e.g., title, hook_sentence, score), which helps prevent the agent from misinterpreting metadata as direct instructions.
Capability inventory: The skill is capable of executing local bash scripts and vendor-specific CLI tools (muapi-cli).
Sanitization: Data sanitization is delegated to the server-side virality framework and transcription engines at the /ai-clipping endpoint.

muapi-youtube-shorts