The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data including text prompts and external resource URLs (images, videos, audio) which are interpolated into API requests. This represents a standard attack surface for indirect prompt injection where malicious instructions could be embedded in the input data.
Ingestion points: User-provided prompt, image_url, video_url, and audio_url parameters in the Python functions defined in SKILL.md.
Boundary markers: None present in the example code snippets.
Capability inventory: Performs network operations via requests.post to external vendor endpoints.
Sanitization: No input sanitization or validation is demonstrated in the provided code examples.
[COMMAND_EXECUTION]: The documentation provides examples for using the modelslab CLI tool to search for and inspect video models. These commands are local to the user's environment and intended for interacting with the vendor's service.
[EXTERNAL_DOWNLOADS]: The skill initiates network requests to modelslab.com to trigger video generation and poll for results. These interactions are directed at the official vendor infrastructure for the purpose of fulfilling the skill's primary function.

modelslab-video-generation