video

What problem does it solve? Creating short video clips with AI generation requires careful prompt design, correct model control selection, and disciplined handling of references, durations, and retries—mistakes waste significant generation credits and produce unusable output. ## Core Features & Use Cases - Structured prompt authoring: Enforces a six-part prompt format covering shot design, camera intent, production detail, sound, dialogue, and ending frame, with pacing matched to clip duration. - Lane and model management: Lists signed-in video lanes and model controls (aspect ratio, resolution, duration ranges, max references) before generating, and respects the user's remembered Studio selection. - Reference-based generation: Supports image-to-video, multi-reference identity preservation, and clip continuation from a previous clip's last frame. - Use Case: A user needs a 6-second product clip for a social feed. The skill generates a still first via the image skill, uses it as a reference, picks 9:16 aspect and a duration from the model's list, generates once at draft resolution to check motion, then delivers at the highest resolution. ## Quick Start Use the video skill to generate a 6-second 9:16 clip of a coffee cup on a wooden table with steam rising, saved as clip.mp4.