gnipbao/story-to-handdrawn-video

What it solves

This project converts text-based stories or sequences of hand-drawn images into vertical (3:4) hand-drawn animations. It simplifies the process of creating stylized, silent video tracks that are ready for post-production voiceover and background music.

How it works

The system consists of a Remotion-based renderer that handles storyboarding, animation, and rendering. It can be driven either via Python scripts or through a "Skill" package that allows AI agents (like Codex, Claude Code, or Kimi Code) to automate the entire pipeline—from sentence splitting and image generation to final rendering—using natural language commands.

Who it’s for

Content creators and storytellers who want to quickly generate hand-drawn style animations from text or existing sketches without manually animating each frame.

Highlights

• 20 Built-in Styles: Includes diverse visual families such as ink wash, watercolor, kid crayon, zine collage, and whiteboard explainers.
• Agentic Workflow: A dedicated skill package allows AI agents to manage the rendering process via natural language.
• Dynamic Visuals: Supports a "text $\rightarrow$ black-and-white sketch $\rightarrow$ colored illustration" reveal effect and optional page-flip transitions.
• Flexible Input: Works with either raw story text or uploaded images, with support for both high-resolution final renders and fast previews.