Repolode
Submit
threerocks

threerocks/sweety-handdraw-video

A Codex and Claude Code skill that turns a topic, article or story into a narrated hand-drawn explainer video, with stroke-by-stroke drawing and Chinese handwriting.

View on GitHub
Stars
761
Forks
140
Open issues
2
Last commit
yesterday

What is it?

sweety-handdraw-video is a skill for AI coding agents that produces narrated explainer videos in which hand-drawn illustrations are drawn on screen. Given a topic, an article or a complete story, the agent writes the narration, plans the storyboard and generates illustrations, while local tools handle line drawing, colour filling, handwritten text, voice alignment and MP4 export.

The pipeline builds the narration first, generates it segment by segment with noise and clipping checks, and then schedules drawing, subtitles and scene changes against the resulting timestamps. Line-art styles are drawn stroke by stroke, while styles without outlines are revealed with a brush. Chinese characters within GB2312 are written in standard stroke order. The project offers 21 built-in style presets, six arm presets and several voice backends, including a local VoxCPM2 adapter and a VoiceStudio HTTP service.

Rendering runs on Python with Pillow, SciPy and FFmpeg. Image generation depends on the host agent or a user-supplied tool, and the project targets macOS and Linux.

Key features

  • 01Narrates the script first, then draws illustrations, filled colour and handwritten text to match the timestamps.
  • 02Draws line-art styles stroke by stroke and reveals other styles with a brush.
  • 03Writes Chinese characters in standard stroke order for characters within GB2312.
  • 04Ships 21 built-in illustration styles and six hand presets, plus custom style descriptions.
  • 05Supports local VoxCPM2 and VoiceStudio voice backends, and reuse of an existing narration.
  • 06Keeps scripts, assets, timestamps and check records so that subtitles can be revised without regenerating images.

Use cases

  • Turning a technical concept or product mechanism into a short explainer video.
  • Adapting a story or essay into a narrated hand-drawn video with a chosen art style.
  • Producing a series of videos that share a fixed style, hand and voice.
  • Revising subtitles or layout of an existing video without redrawing every illustration.

Topics

#video-generation#ai-agent#text-to-speech#illustration#explainer-video#python#ffmpeg#claude-code

FAQ

What is sweety-handdraw-video?
A Codex and Claude Code skill that turns a topic, article or story into a narrated hand-drawn explainer video, with stroke-by-stroke drawing and Chinese handwriting. sweety-handdraw-video is a skill for AI coding agents that produces narrated explainer videos in which hand-drawn illustrations are drawn on screen. Given a topic, an article or a complete story, the agent writes the narration, plans the storyboard and generates illustrations, while local tools handle line drawing, colour filling, handwritten text, voice alignment and MP4 export.
Is sweety-handdraw-video free to use? What license does it have?
sweety-handdraw-video is open source under the MIT license. Check the license terms before using it in commercial products.
Is sweety-handdraw-video still maintained?
Actively maintained. The last commit to sweety-handdraw-video was on Oct 19, 2026.
What can I use sweety-handdraw-video for?
Turning a technical concept or product mechanism into a short explainer video. Adapting a story or essay into a narrated hand-drawn video with a chosen art style. Producing a series of videos that share a fixed style, hand and voice. Revising subtitles or layout of an existing video without redrawing every illustration.
Where can I get sweety-handdraw-video?
The source code and installation instructions are on GitHub: https://github.com/threerocks/sweety-handdraw-video

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects