The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the belt CLI tool to interact with the video generation models. It implements security best practices by using the allowed-tools frontmatter field to restrict the agent's shell access exclusively to the belt command.
[EXTERNAL_DOWNLOADS]: The skill references documentation and installation instructions hosted on official inference.sh domains and their public GitHub repositories. These are standard resources for the service provided by the skill.
[PROMPT_INJECTION]: As a generative AI utility, the skill processes user-supplied prompts for video creation. This represents a standard surface for indirect prompt injection where a user might attempt to bypass generation filters, though this is managed by the underlying model provider's safety systems.

google-veo