Fusion motion graphics

Make the message clear in the settled frame, then use motion to direct attention. Keep type, shapes, masks, and animation editable in Fusion.

Compose the scene

Decide the focal phrase, supporting visual, palette, type hierarchy, and delivery format. Build a readable hero frame before adding entrances. Use the user's reference to measure line breaks, relative size, placement, color, and the timing of each visible event. A template is a starting point; its default style is not a match to the reference.

Read references/design-and-layout.md for typography, cards, diagrams, or reference matching. Prefer fewer well-composed elements to a larger graph with no reading order. Check at the intended viewing size: thin strokes, crowded text, and soft contrast often disappear on a phone.

Choreograph attention

Plan entrance, reading hold, and exit around the actual copy. Reveal supporting geometry before its label when that establishes meaning. Animate related objects together; stagger only when the order communicates something. Use a restrained settle for editorial work, an overshoot for an intentionally elastic object, and constant motion for a mechanical sweep.

Read references/animation-and-timing.md when choosing choreography or adapting a reference's rhythm. Give the viewer time to read before beginning the next competing movement. Match the entire movement, not just its final frame.

Build with the installed SDK

For local execution, read cutagent-fusion; check the installed SDK declarations, public CutAgent CLI command cards, and live capabilities for exact operations.