项目的 README,来自 github.com/kaomei/stickman-video-director。合并为一个 markdown 文件:README.md。
🇺🇸 EN · 🇨🇳 简中 · 🇯🇵 日本語 · 🇰🇷 한국어 · 🇫🇷 FR · 🇧🇷 PT-BR · 🇹🇷 TR · 🇮🇩 ID
Turn any script into viral, multi-style animated stickman videos for Faceless YouTube Channels.
Start with a topic, an article, or a script. Automatically orchestrate visuals, voiceovers, captions, and sound effects into a production-ready stickman video.
⚡ 100% Free & Open-Source. No camera, no $50/mo subscriptions, and ready in minutes, not days.
Continually updated visual styles for science explainers, dark psychology, tech parables, viral lore, and storytelling channels.
Classic Black & White
|
Modern Pop
|
Prehistoric Cave
|
Cozy Storytelling
|
Historical Cartoon
|
Minimal Line Art
|
Amber Noir
|
Colorful Hand-Drawn
|
Retro TV
|
Survival Thriller
|
Character Explainer
|
Coming Soon… |
11 visual styles · Tested examples · Free and open source
⭐ Star the project to catch the next style drop and feature unlocks.
Get started · Styles & Demos · FAQ · Roadmap · Community
Want control over every shot? Choose A · Director Mode. Want a finished video without assembling scenes yourself? Choose B · Autopilot Mode. One project, two workflows.
A · Director Mode |
B · Autopilot Mode |
|
|---|---|---|
| Approach | You control production; AI plans the shots. Get storyboards and prompts, then generate clips and edit them yourself. | Give it a topic or a brief; get a finished video. Automate illustrations, voiceover, captions, subtle animation, and video assembly. |
| Length | Short-first, typically 30 seconds–3 minutes. Longer videos can be built scene by scene. | Long-form-first, especially narrated videos of 10 minutes or more. Also works for shorts and episodic stories. |
| Tools | Plan with a Skill-compatible AI assistant, generate clips in your preferred video generator (Google Flow, Runway, Luma, Kling, etc.), then assemble in CapCut or Premiere. | Run in Codex, Claude Code, or another AI coding agent that can read and write files, call tools, and execute commands to connect generation with local rendering. |
| Visuals | Generated animated clips, with character movement, camera direction, and interactions. | Illustrated scenes with local animation: expressions, character reactions, and prop movement support the narration alongside cuts, pans, and zooms. Camera movement or text entrances alone do not count as picture animation. |
| Output | A director's treatment, storyboards, scene-by-scene video prompts, and sound design. | A complete video file with voiceover, captions, and subtle animation. |
Both workflows can start with a raw topic, a web link, or a full script. Customize your aspect ratio (9:16, 16:9, 1:1), narration language, and voiceover freely. The lengths above are recommendations, not hard limits.
💡 Quick Setup Note:
![]() 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. |
For results closest to the showcase, I recommend starting with Codex. I use Codex to create and test my showcase examples for this project.
You can also install the Skill in Claude Code, Cursor, Antigravity, or other agents with file and terminal access. However, differences in models, tool configuration, and execution can lead to different results.
Send the following to your AI assistant:
Install this Skill for me:
https://github.com/kaomei/stickman-video-director
Once installed, read the installed SKILL.md right away.
Explain what each mode produces, the tools it needs, and the workflow
in plain language, with examples I can copy.
Based on what I want to make, tell me what is missing and exactly what to do next.
After installation, send this to your AI assistant and replace the placeholder with your topic:
Use Director Mode / mode A to turn the following topic, copy, or brief
into a stickman video:
[Enter a topic, paste your copy, or describe what you need]
Requirements: landscape 16:9, A3 "Prehistoric Cave" style, starting with 30 seconds.
Give me the director's treatment first. After I approve it, provide scene-by-scene video prompts.
Workflow: share your topic and brief → develop the story script → optionally approve front, side, and back character views → generate detailed storyboard prompts → manually generate clips and edit the final video
The treatment covers the story structure, scene-by-scene narration, actions, camera direction, and sound design. Copy each prompt into Google Flow or another video generator, then add captions and finish the edit in CapCut or your preferred editor.
For your first video, follow the visual production walkthrough on X (Thread with step-by-step video demos).
After installing in Codex, Claude Code, or another environment with file access and command execution, try:
Use Autopilot Mode / mode B to turn the following topic, article, post, news item, or brief
into a compelling YouTube video:
[Paste content or a link, or describe what you want to promote]
Requirements: 5–10 minutes, landscape 16:9, B1 "Minimal Line Art" style, with voiceover, captions, and sound effects.
Check the image generation, voiceover, and video-rendering setup first. Then give me a script and a 20-second preview with sound.
After I approve it, build the full video with visuals timed to the narration. Deliver a complete video file; I do not want to assemble scenes myself.
Workflow: share a topic, content, or brief → approve the script → approve a 20-second preview with sound → automatically produce the full video → receive the finished video file
The AI assistant plans the scenes, generates the visuals and narration, adds captions and sound effects, and renders a complete video to the voiceover's timing. You approve the script and preview; you do not have to generate and assemble every scene yourself.
On first use, it checks the required setup and walks you through anything missing. If a visual or voiceover section needs work, revise just that section and keep what you have already approved.
Optional finishing for either mode: once you confirm the final video, continue with the Thumbnail & Publishing Kit for a matching cover and upload copy—or finish here. No new mode or installation is needed.
Styles are numbered A1, A2, A3… Each includes its own animated examples.
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.
| A1.1 · Black on white | A1.2 · White on black |
|---|---|
![]() |
![]() |
| Pure white canvas · Black characters | Pure black canvas · White characters |
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.
| A2.1 · Minimal Tech | A2.2 · Colorful Story |
|---|---|
![]() |
![]() |
| White space · Faint grid · Cyan glow | Colorful scenes · Cinematic lighting · Story atmosphere |
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.
| Example 1: "How Did Humans Get Their First Job?" | Example 2: "The First Discovery of Fire" |
|---|---|
![]() |
![]() |
| Prehistoric story · Workplace twist | Discovery of fire · Community creation |
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.
| Example 1: Message anxiety on the commute | Example 2: Making time for yourself |
|---|---|
![]() |
![]() |
| Subway station · Walking shots · Emotional close-ups | Outdoor café · Warm light · Everyday gestures |
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.
| Example 1: Edo Japan · A samurai's everyday life | Example 2: Ancient Egypt · Cats and mummies |
|---|---|
![]() |
![]() |
| Edo streets · Samurai clothing · Everyday contrasts | Colorful temples · Black cat characters · History with a twist |
Style and prompt contribution: Lawal Ridwan / @leksautomate.
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.
| Example 1: What if everything cost 10,000 times less? | Example 2: The Era of Collapsing Prices |
|---|---|
![]() |
![]() |
| Renoise MCP example · 20 seconds · Narrated by Tim | Google Flow example · 8 seconds · Model-native audio |
New specialized director styles are currently in training. Star the project to unlock them at 5,000 Stars!
Styles are numbered B1, B2, B3… Each has its own finished examples combining illustrated scenes with subtle animation, maintained separately from the A series.
Like a friend sketching an idea as they explain it. Expressive faces, black linework, and a few colorful props turn abstract knowledge into little scenes—and product benefits into situations people recognize.
Why does your mind get louder the harder you try to sleep? Inspired by YouTube's @PsychToonsHQ (200K subscribers), warm paper textures, soft colors, and handwritten touches turn sleeplessness, procrastination, and distraction into little stories that feel uncomfortably familiar.
Like stumbling across a curious late-night broadcast. Inspired by YouTube's @BradikininVideos (55K subscribers gained in 3 weeks), rounded white-and-gray characters, faint bloom, and analog texture turn an ordinary question into a satisfying discovery.
A countdown. A late-night call. A rule you must never break. Ordinary places become traps in a look inspired by YouTube's @BeanieSurvives (100K subscribers gained in 3 weeks). Round white heads, white gloves, and stark light and shadow give simple characters an unsettling world to inhabit.
Give your favorite avatar a starring role—permission is required to use well-known characters. Your presenter takes the mic in the corner while expressive stickmen act out the idea—dragged around by bills, flattened by pressure, or caught in a visual punchline. Even abstract ideas become something viewers can see and remember.
Your video is only as good as the click. Once your animation is complete, orchestrate matched thumbnail art, high-CTR titles, and viewer engagement hooks in the same workflow.
An integrated finishing pack for both Director and Autopilot Mode: generates visual hooks tailored to your video's approved character and art style, complete with ready-to-publish YouTube metadata.
| A simple question and one clear symbol. | A story setup with an oversized obstacle. | Period costume, scenery and display lettering. |
| One conflict, told through extreme scale. | Busy versus effective, at a glance. | A relatable scene, painted light and hand lettering. |
All thumbnail concepts are designed around proven YouTube packaging psychology: bold visual contradictions, clean silhouettes, and mobile-readable typography.
After you confirm the final video, your assistant will ask whether you want this optional finishing pack. Say yes to continue, or skip it without delaying video delivery. No need to copy a prompt. You can also ask directly:
Generate the Thumbnail & Publishing Kit for this video:
- 1 landscape 16:9 thumbnail image matching the approved art style
- 5 YouTube title options
- A description, relevant tags, and verified timestamps if available
- 1 engagement question for the pinned comment
- A pre-publish checklist with any pending checks clearly marked
Image generation requires an available tool and any necessary cost approval; if unavailable, the assistant explains the gap and still delivers the text pack. An early request based on an approved script produces drafts, with final-cut checks pending. Download the cover and copy the text into YouTube Studio or TikTok as appropriate; nothing is uploaded or published automatically.
Yes. Stickman Video Director is 100% free and open-source under the MIT license. You can use it freely for both personal and commercial projects. Unlike proprietary SaaS platforms that charge $30–$60/month, this project has zero subscription fees. You only pay for your own third-party API usage (e.g., Google Gemini, Claude, or ElevenLabs) based on the models you choose to connect.
B · Autopilot Mode: Uses a code-driven production workflow. Character image assets are created before the video is produced and reused across scenes, so character consistency is relatively stable.
A · Director Mode: Primarily relies on text-to-video generation. If a character has a complex hairstyle, outfit, or other features that are difficult to reproduce consistently from text alone, add a stick-figure character reference image and keep using it for subsequent clips to make the results more consistent.
I recommend Codex for results closest to the examples. I created and tested the B Mode showcase videos in Codex; you do not need to write code yourself. Other agents can install the Skill, but differences in models, tool configuration, and execution can affect the results.
My examples use GPT's built-in model without a separate image-generation API. Other agents should use their available image tools; B5 real-person or complex hosts need generated or approved image assets, not code-drawn substitutes. Voice credits describe the examples, not a required provider. Ask your agent to configure your preferred service, then check a short sample before making the full video. Style presets are defaults, not restrictions: ask your agent to change fonts, text effects, or the video language and adapt the layout to your request.
Yes. You can paste a rough idea, a shower thought, a news link, or an entire blog post. The AI director will extract the core conflict, write a compelling 3-second hook, and structure the story and shots to fit your topic, target length, and chosen style, complete with comedic timing and voiceover cues.
Traditional AI tools just spit out literal illustrations of text. Stickman Video Director applies real visual storytelling psychology:
Yes, absolutely. A major concern for faceless creators is YouTube's "Reused Content" or "Low-effort AI" demonetization policy. Because Stickman Video Director writes custom storytelling arcs, unique narrative pacing, and original visual staging, your output is completely original. Hundreds of channels successfully monetize stickman narrative videos under the YouTube Partner Program (YPP).
No. You do not need a high-end RTX graphics card or coding knowledge.
Every star fuels the next drop! Together, we’re building a 100% free, open-source studio to help solo creators run faceless channels without $50/mo SaaS subscriptions.
We are actively training new specialized engines for the most requested faceless YouTube niches:
Got a style idea or want to get featured? Join our Telegram Community:
🔔 Follow @CrazyKaomei on X for daily video workflows, automation tools, and channel growth SOPs.
Released under the MIT License. Free to use, modify, distribute, and use commercially. Keep the original copyright and license notice when distributing the project or substantial portions of it.
Provided “as is,” without warranty. See LICENSE for the full terms.