The Agent Skills Directory

[SAFE]: The skill implements a standard workflow for media transformation using a designated API (MuAPI) with clearly defined capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided descriptions and prompts to drive video transformations.
Ingestion points: User-supplied source URLs/paths, effect intent, and descriptive prompts are processed at runtime.
Boundary markers: The instructions include a 'Safety and scope gate' and require specific descriptive structures, though explicit technical delimiters for prompt interpolation are not defined.
Capability inventory: The skill uses media.edit_video, media.upload_file, and media.check_result. It does not have shell access or arbitrary file-system write permissions.
Sanitization: The skill relies on 'Safety and scope gate' logic to filter harmful content before submission to the API.
[DATA_EXPOSURE]: The skill utilizes the media.account_balance capability to monitor API usage costs, which is a standard functional requirement for professional media processing workflows.

Video Effects