kaomei/stickman-video-director
Turn any script into viral, multi-style animated stickman videos for Faceless YouTube Channels. Automate visuals, voiceover, captions & SFX. 100% Free & Open-Source.
About kaomei/stickman-video-director
kaomei/stickman-video-director is an open-source project on GitHub, mainly written in JavaScript. Turn any script into viral, multi-style animated stickman videos for Faceless YouTube Channels. Automate visuals, voiceover, captions & SFX. It currently holds 5,348 stars and 869 forks with 3 open issues, and was last pushed on 2026-11-10 (repository created 2026-08-03).
Project Overview
Git Homed tracks it on the Video Trending board and on the AI Video Trending list.
GitHub Repository Details
README
🇺🇸 EN · 🇨🇳 简中 · 🇯🇵 日本語 · 🇰🇷 한국어 · 🇫🇷 FR · 🇧🇷 PT-BR · 🇹🇷 TR · 🇮🇩 ID
🎬 Stickman Video Director
Scale your faceless channel with automated stickman animations.
Turn raw scripts into high-retention animated explainers in minutes — across 14 battle-tested styles, complete with voiceovers, dynamic camera moves, and sound design.
No After Effects. No $50/mo SaaS tax. Ship daily without burnout.
Quickstart · Showcase · FAQ · Roadmap · Community
⭐ Star the project to catch the next style drop and feature unlocks.
🎬 Two video creation modes: Model Engine vs Code Engine
One stickman director, two ways to create your visuals: A uses video models to generate continuous shots; B uses code to animate layered assets. With the right tools, an Agent can take either approach through to a finished video, following the selected style’s specifications.
A · Model Engine |
B · Code Engine |
|
|---|---|---|
| Concept | Feeds prompts and storyboards into video generation models to render fluid video clips directly. | Generates layered visual assets, then writes and executes code to animate, compose, and render locally. |
| Features | Fluid natural motion and cinematic lighting; vivid visuals with rich cinematic dynamics. | Stable character consistency with reusable assets; precise timeline pacing, layout control, and easy local fine-tuning. |
| Best for | Creators with access to video generation resources (e.g., Google Flow, affordable APIs, or self-hosted local models), focusing on visually striking short-form content. | Users with active Codex or Claude Code subscriptions looking to put surplus coding quota to work—without paying extra for expensive video model credits. |
| Workflow | Confirm script and style to get scene prompts for manual generation; or let the AI operate a built-in browser to generate clips, stitch scenes, and handle subtitles & audio. | Confirm the video script and a 20s sample for the chosen style, then let the AI automatically handle all asset preparation, code authoring, animation execution, subtitles, audio, and local rendering. |
| Output | Storyboard script + scene prompts + (full video file) | Layered asset library (characters/scenes/props) + full video file |
| Cost | Consumes video model generation credits/points; longer videos scale significantly in cost and re-roll overhead. | Zero video model credits needed; renders locally using your existing Agent quota and basic image/TTS tools. |
Start with an idea, an article, a web link, or a complete script. Both modes support short videos and longer content. Choose 16:9 landscape or 9:16 portrait for your publishing platform, and configure subtitles, voiceover, background music, and sound effects to suit your needs and the selected style.
Still deciding on a style? Below, you’ll find detailed descriptions, suggested topics, and more sample videos for each style.
API Partners · Support this project ↗
![]() SJolt |
SJolt brings image, video, and audio generation models together through one API. Compare output quality and cost in the playground, then bring the same request flow into your project—from experimentation to integration, without switching tools. Track asynchronous tasks and pay by usage, with one balance across supported models. Explore SJolt to browse models and try the playground. |
![]() |
APIMart is a low-cost API platform for AI image & video generation — GPT-Image-2 from $0.006/image, 160+ images per dollar. One async API covers both image and video: submit a task, get an ID, fetch results via polling or callback. Batch tens of thousands of images without timeouts, switch models without changing code. Pay-as-you-go with no monthly fee — sign up here to get started. |
![]() Renoise |
Renoise AI is an all-in-one AI video and image creation platform for marketers, advertisers, and content creators. Tell the Agent what you need in plain language—it writes prompts, breaks your idea into shots, and generates videos in batches, keeping the same character or product consistent across shots. Create directly on the web or in the iOS and Android apps, or connect Renoise to Claude or ChatGPT (Codex) via MCP. Generate assets, edit, and polish your video right in the conversation—no editing software required—and get a finished video ready to publish. Visit Renoise to explore its models and creative tools. |
🧭 Get started
For results as close as possible to the sample videos, we recommend AI tools with strong coding capabilities, such as Codex or Claude Code.
You can also use this skill with other agents that can read and write files and run commands. Results may vary with the model’s capabilities, tool setup, and execution approach.
Step 1: Install the skill
On first use, the AI will check the required setup and help you install or configure anything missing. Send the following to your AI assistant:
Help me install this skill:
https://github.com/kaomei/stickman-video-director
After installation, read SKILL.md and tell me you’re ready.
Step 2: Choose a style and mode, then start with a prompt
Choose a style you like from A Mode or B Mode, then copy the corresponding prompt below, adapt it to your needs, and send it to your AI assistant:
A · Model Engine
Use Model Engine (A Mode) to turn the following topic or text into a stickman video:
[Enter a topic, paste your text, or describe your idea]
Requirements: 16:9 landscape, style A3 “Prehistoric Cave.” Start with a 30-second video.
First, give me a story outline, an audio plan, and a storyboard.
After I approve them, provide self-contained video generation prompts for each scene.
- Production workflow: After approving the storyboard, you can generate the clips manually or ask your agent to generate them using its browser tools. Start with a low-resolution test. Once you approve the result, your agent can handle the remaining clips and assemble the full video, adding subtitles and audio as needed.
- Recommended platform: Use Google Flow to get closer to the sample results. For your first video, see the detailed tutorial on X.
B · Code Engine
Use Code Engine (B Mode) to turn the following topic or article into an engaging stickman video:
[Paste an article, a news story, or a link, or describe your topic]
Requirements: 16:9 landscape, 8 minutes, style B1 “Minimal Line Art,” with voiceover, subtitles, and sound effects.
First, check that the local tools for image generation and rendering are set up. Then provide a complete script and a preview of the opening 20 seconds with audio.
After I approve both, automatically prepare all assets, write the code, and render the complete video locally.
- Production workflow: After you approve the script and the preview of the opening 20 seconds with audio, the agent prepares the assets, animation, subtitles, and audio to match the selected style, then renders the full video locally and provides a preview.
- Targeted edits: If you’re unhappy with part of the visuals or audio, ask the AI to adjust just that section while keeping the approved parts.
Step 3: Create a thumbnail, title, and description
Once you’re happy with the finished video, continue with the Thumbnail & Publishing Kit. Send this to the AI:
Create a thumbnail, video title, description, and pinned comment for the video we just made, using the approved visual style.
The AI will generate a matching thumbnail, video title, description, and pinned comment in one step, without switching modes or reconfiguring your setup.
🎨 A · Model Engine Overview
The current A Mode samples were created with Google Flow. You can also choose another video generation platform. Before asking an Agent to produce a video automatically, make sure you’re logged in and have available credits.
Styles are numbered A1, A2, A3… Each style includes its own animated samples.
A1: Classic Black & White
A round head and a handful of lines can make a complicated idea click. With the clutter stripped away, every gesture and visual metaphor matters: anxiety becomes a cage; attention becomes a battery running low.
- The look: Crisp black ink on white, or a more dramatic white-on-black version. A few bright accents put the important detail in focus.
- Made for: Science, psychology, personal growth, business breakdowns, and concepts that become clearer with the right visual analogy.
|
|
| Pure white canvas · Black characters | Pure black canvas · White characters |
---
A2: Modern Pop
Red beanie. Yellow shirt. A character you recognize at a glance. Follow them from floating tech interfaces into city streets and the wilderness, turning a tool demo or a personal turning point into a character-led story.
- The look: Minimal Tech pairs white space with glass-like interfaces and cyan light. Colorful Story adds richer settings, depth, and mood-driven lighting.
- Made for: AI tools, tech products, tutorials, personal growth, everyday experiences, and adventures.
|
|
| White space · Faint grid · Cyan glow | Colorful scenes · Cinematic lighting · Story atmosphere |
---
A3: Prehistoric Cave
No phones. No offices. So why are the cavemen still stuck working? Visually inspired by YouTube’s @Zenn0009 (200K subscribers). Drop modern problems into a world of caves, campfires, and animal skins, and watch familiar frustrations turn into absurd little stories with a sting.
- The look: Rough hand-drawn lines, earthy settings, and big reactions. Character interactions, reaction shots, and comic pauses do the storytelling.
- Made for: Workplace satire, human nature, prehistoric what-ifs, dark comedy, fables, and English-learning stories for children.
|
|
| Prehistoric story · Workplace twist | Discovery of fire · Community creation |
---
A4: Cozy Storytelling
The message you feel you must answer. The afternoon you finally keep for yourself. Small moments in stations, bedrooms, and corner cafés give hard-to-name feelings a place to live—and viewers a chance to recognize themselves.
- The look: Colorful everyday settings, gentle light, and understated gestures. A close-up catches hesitation; a wider shot leaves room to breathe.
- Made for: Relationships, everyday emotions, personal growth, habits, and warm stories that still make room for anxiety and exhaustion.
|
|
| Subway station · Walking shots · Emotional close-ups | Outdoor café · Warm light · Everyday gestures |
---
A5: Historical Cartoon
Samurai have bills to pay. An Egyptian cat can steal the scene. Step into the streets, temples, and camps of the past, turning a historical detail into something a character actually has to deal with.
- The look: White round heads, bold ink outlines, and colorful period settings. Costumes, buildings, and props establish the era; reactions and playful contrasts bring it to life.
- Made for: World history, curious civilizations, everyday life in the past, historical figures, and humorous history shorts.
|
|
| Edo streets · Samurai clothing · Everyday contrasts | Colorful temples · Black cat characters · History with a twist |
Style and prompt contribution: Lawal Ridwan / @leksautomate.
---
A6: Amber Noir
What if you woke up and everything cost ten thousand times less? The visual inspiration is YouTube's @IfverseTales (60K subscribers gained in 3 weeks). An orange-hooded figure moves through golden light and deep plum shadows, turning a casual hypothetical into a cinematic story about temptation, consequences, and unexpected turns.
- The look: Amber highlights, dark silhouettes, and layered city scenes. Character actions and restrained camera movement build curiosity without a talking presenter.
- Made for: What-if scenarios, modern fables, psychology, unexpected discoveries, and stories with a twist.
| 










