A mode of the merged Skyreels/VACE workflow (Inner Reflections' Skyreels + VACE merged model) that combines several reference images, such as characters and an environment, into one video. A natural-language prompt says who is where and what they do. It replaces the single start-frame mode described in Single start-frame vs multiple references for style guidance.
The demo scene was built simply. The author downloaded Mixamo animations into Blender and rendered an outline pass (Outline pass in Blender (Workbench + Freestyle)). A rough layout with stock animation is enough input.
If a character's details change between shots (the demo fox's outfit changed), make the prompt more specific about that detail. See LLM-assisted prompt refinement and restyling for getting help writing it.
The frame load cap of 81 is the single-iteration sweet spot. Longer shots need chaining; see overlapping-start-frame-iterative-batching.