The Working Approach
motion-canvas-agent connects an agent to a running Motion Canvas editor through a local HTTP API. It can seek to a frame, capture a screenshot, inspect the scene graph, change preview or render settings and trigger an export. The browser still runs the player and renderer.
The integration uses two files: a Vite server plugin and a client plugin inside the Motion Canvas project. Requests travel through the development server to the open browser. This makes exact-frame inspection scriptable, but it requires a live editor connection rather than acting as an independent headless renderer.
Expected Deliverables
- A local control bridge attached to a Motion Canvas project.
- Frame-specific screenshots and scene-graph inspection data.
- Requested render actions and recorded preview/render settings.
From Brief to Review
- 01Connect both plugin sides
Add the server plugin to Vite and the client plugin to the Motion Canvas project. Start the development server and open the editor in a browser.
- 02Inspect a known frame
Check connection status, set the intended size and seek to an exact frame. Capture it and compare scene-graph values with the visible result.
- 03Render with explicit settings
Inspect errors and distinguish preview settings from rendering settings. Trigger the requested export and verify its actual artifact after completion.