GPT Image 2 keyframes wired into Seedance shots on the board.
Skill ID gpt-image-2-seedance-2-workflow
Example outputUnder the hood
You write a line. The agent does the rest — it studies this skill's guide, expands your idea into a full production brief, and runs the tools itself.
All 8 parts of the guide — What this skill does, 1. The tools this skill uses and more.
Turns your one line into a detailed, in-style prompt — about as rich as the 95-word example.
Generate image, Generate video, Read canvas, Edit canvas — chosen and called by the agent, not by you.
Your clips land in Create, ready to edit, remix or animate.
Production storyboard-to-video workflow using GPT Image 2 + Seedance. Use ONLY for multi-scene, long-form, or storyboard-based video production — NOT for simple single image or single short video generation.
The 8-part playbook the agent reads before it generates.
The exact brief that produced the image above — the level of detail the agent writes for you from a single line.
Flat overhead view on a bone-white studio surface, filling the frame. Four square keyframe cards in a horizontal row, each showing the same bright sunlit coastal path at midday, connected left to right by thin warm grey connector lines with small round nodes, one node picked out in amber. Beneath the row, one wide letterboxed film frame of the same bright scene with sprocket holes along each edge. Soft even daylight from above, delicate drop shadows under each card, bone-white and pale sand palette with a single amber accent, bright and high key, no lettering.