A frame-capture workflow advances the animation through known times and saves the visual result for each frame. Deterministic capture avoids screen-recording timing variation and can retain transparent pixels when the renderer and image format support alpha.
Reliable capture checks that every expected frame exists, uses consistent dimensions and records the intended frame rate. Dynamic pages may also need fonts, assets and animation state to finish loading before each image is saved.
ELI5
Frame capture saves one picture for every step of an animation. The pictures can then be combined into a video with exact timing.
For example, an agent capturing at 30 frames per second saves 30 images for each second of motion. It must wait for the correct animation state and verify that no frame was skipped or duplicated.
