What problem does it solve?
This skill converts structured cinematic or storyboard-style prompts and optional reference images into runnable video generation jobs and final MP4 outputs, removing the need to manually craft API requests, manage uploads, or handle long-running polling and downloads.
Core Features & Use Cases
- Structured JSON prompts capture title, scene description, characters, camera movement, dialogue and audio cues to produce consistent, reproducible generation inputs.
- Reference image guidance accepts one or more images to define visual style or key frames and can act as a guided first/last frame.
- Automated Python execution runs a provided script that calls the Gemini long-running video generation endpoint, polls for completion, and downloads the generated video into the outputs directory.
- Use Case: Quickly prototype a short scene from a storyboard, create promotional clips from a narrative brief, or generate visual drafts for iterative creative review.
Quick Start
Generate a 16:9 short clip of a rainy city street at dusk by saving a structured JSON prompt to /mnt/user-data/workspace and running the skill to produce /mnt/user-data/outputs/rainy-street.mp4.
