The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to pass user-provided prompt text directly into the belt CLI tool for video generation. This creates a surface where maliciously crafted prompts could attempt to influence the agent's behavior. * Ingestion points: Prompt values in the JSON input for belt app run commands in SKILL.md. * Boundary markers: None specified in the instructions to separate data from instructions. * Capability inventory: Execution of the belt tool via Bash (restricted by frontmatter to the belt command). * Sanitization: No explicit sanitization or validation steps are defined for the input prompts.

google-veo