Series E · A Model Reasons, The Objects Move

The Commandable Scene

SVG + Canvas · type a command · a model turns it into actions

Type a plain-language command. A model interprets it into a structured action, and the action manipulates the objects — in both an SVG layer (clickable shapes) and a canvas layer (animated points). Two engines: the live Claude API when this runs inside the Claude.ai artifact runtime, and a local interpreter everywhere else. Same actions, either way.

◆ SVG layer · structured shapes

◆ Canvas layer · animated points

objects: 0 purple 0 blue 0 green 0 rose 0
ready — type a command and the model will move the objects
two engines, one action schema: the command goes to a model that returns a structured action like {op:"add",color:"#3a6a87",n:6}. Inside the Claude.ai artifact runtime, that's the live Claude API (claude-sonnet-4-6) actually reasoning about your words. Standalone (like opened from a file), it falls back to a local rule interpreter that returns the same shape — so it always works, and lights up the live path when the runtime provides it.
TYPE A COMMAND → A MODEL RETURNS A STRUCTURED ACTION → THE OBJECTS MOVE (SVG + CANVAS)
LIVE CLAUDE API IN THE ARTIFACT RUNTIME · LOCAL INTERPRETER EVERYWHERE ELSE · SAME ACTION SCHEMA
THE COMMANDABLE SCENE · A PURPLE PAPER · SERIES E · JUNE 2026