whiteboard-mask-animation

whiteboard-mask-animation is an Agent Skill published by GitHub user geeklee for whiteboard explainer animation. The README says it turns static line drawings from a Chinese article or storyboard into an editable, previewable mask animation, exported as MP4 once you approve it. The AI assistant you use reads the article and proposes an illustration plan. After you confirm, it generates 16:9 line drawings and marks regions in each picture according to the text. The order, direction and timing of each region live in an annotation.json file named after the image. You check the animation in a local browser preview first, then export it. The project does not name a model or image service. The same author's SRT Whiteboard Animation is already listed on this site and combines this project's mask reveal with the continuous strokes of whiteboard-stream-animation. This one starts from the article and decides when each part of the picture appears.

Who it's for:Content creators who want to pair an article with pictures and turn it into an explainer that reveals one section at a time.

Before you start: The README says to put it in your assistant's Agent Skills directory or reference it directly in Codex; support for other assistants is not stated. You need Python 3 and the Pillow, NumPy and OpenCV packages installed with pip. The repository includes a preview launcher for Windows; full operating system support is not stated. The project does not say whether a paid API or key is needed, or whether anything is uploaded when illustrations are generated. The actual image service and its cost depend on the assistant you use. The code is under the MIT license, so commercial use is allowed if you keep the license notice. Example images and anything you feed in carry their own rights.

Based on the project's public documentation. Not tested by AI Inspiration World.