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.
原始仓库简介 · 中文摘要待补充 · agent-skills-manager / ai-animation-generator / ai-video-production / automated-video / codex-skills / content-creation
README / 项目原文
🇺🇸 EN · 🇨🇳 简中 · 🇯🇵 日本語 · 🇰🇷 한국어 · 🇫🇷 FR · 🇧🇷 PT-BR · 🇹🇷 TR · 🇮🇩 ID
AI Stickman Video Generator for Faceless YouTube Channels
Turn any idea into a high-quality stickman video with a consistent visual style.
Start with a topic, an article, or a link. Automatically create visuals, voiceover, captions, sound effects, and subtle animation, and get a complete stickman video. Build a distinctive YouTube channel without ever appearing on camera.
An open-source collection of popular stickman animation styles from across the web, continually updated for science and knowledge explainers, psychology and motivation, education and awareness, uplifting stories, trending stories, and more.
Classic Black & White
|
Modern Pop
|
Prehistoric Cave
|
Cozy Storytelling
|
Historical Cartoon
|
Minimal Hand-Drawn
|
6 styles · 8 visual options · Tested examples · Free and open source
⭐ Star the project to keep up with new stickman styles as they arrive.
Get started · A styles · B styles · FAQ
Two ways to make your video
Want control over every shot? Choose A · Director Mode. Want a finished long-form video without assembling every scene? 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 Google Flow or a similar service, then edit in CapCut or your preferred editor. | 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 subtle animation, using cuts, pans, and zooms to support the narration rather than generating a full-motion clip for every scene. |
| Output | A director's treatment, storyboards, scene-by-scene video prompts, and sound design. | A complete video file with voiceover, captions, and subtle animation. |
| Start |
Both workflows can start with a topic, copy, or a complete script. A and B have separate style numbers, galleries, and visual assets; new styles and examples are added to their own series. Choose the aspect ratio, narration language, and voice you need. The lengths above are recommendations, not limits.
Autopilot Mode needs working image and voice generation services plus a local video-rendering environment. A standard chat window alone cannot replace that setup. The Skill is free and open source; connected services may charge separately.
Get started
Send the following to an AI assistant that supports Skills. After installation, it will explain what each mode produces and how to begin, with examples you can copy.
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.
A · Director Mode: scene-by-scene video prompts
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 full production walkthrough (in Chinese).
B · Autopilot Mode: from a brief to a long-form video
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 Hand-Drawn" 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.
A · Director Mode style gallery
Styles are numbered A1, A2, A3… Each includes its own animated examples.
A1: Classic Black & White
Explain an idea with just a few lines. On a pure white or black canvas, characters, props, and a little accent color do the heavy lifting—ideal for brisk, clear explainers.
- Visual style: simple round heads, line-drawn limbs, and consistent stroke widths, with occasional simple expressions. Up to three saturated accent colors signal emotion, risk, or action. Abstract ideas become visual metaphors: anxiety as a cage, attention as a progress bar.
- Best for: science explainers, psychology, motivation, self-management, business breakdowns, and abstract concepts.
- Choose your look: black characters on white feel clean and direct for teaching and everyday explainers. White characters on black create stronger contrast for suspense, conflict, and emotion.
| A1.1 · Black on white | A1.2 · White on black |
|---|---|
 |
|
| Pure white canvas · Black characters | Pure black canvas · White characters |
A2: Modern Pop
A red beanie and yellow T-shirt give this character an instantly recognizable look. Explain a product in a clean tech setting, or follow the character into bedrooms, city streets, and the wilderness for a story with real ups and downs.
- Visual style: a minimalist character with a consistent red beanie and yellow shirt, in two types of setting. Minimal Tech pairs white space and a faint grid with floating glass-like interfaces and cyan light. Colorful Story uses richer scenery, lighting, and depth to set the mood.
- Best for: Minimal Tech suits AI tools, tech products, tutorials, and feature walkthroughs. Colorful Story works for personal growth, everyday life, adventure, and motivational stories.
- Choose your look: explaining a tool or method? Pick Minimal Tech. Taking viewers through a character's experience? Pick Colorful Story.
| A2.1 · Minimal Tech | A2.2 · Colorful Story |
|---|---|
|
 |
| White space · Faint grid · Cyan glow | Colorful scenes · Cinematic lighting · Story atmosphere |
A3: Prehistoric Cave
Take today's problems back to a world without phones or offices. Caves, animal skins, campfires, and exaggerated expressions turn the absurdities of work and human nature into short stories with a twist.
- Visual style: bold hand-drawn strokes, solid white heads, dark limbs, animal-skin clothing, and earthy surroundings. Character interactions, reaction shots, and comic pauses drive the story. Design the cast for each story—different ages and genders are welcome; not every story needs the same face.
- Best for: prehistoric what-ifs, workplace satire, observations on human nature, dark comedy, fables, and English-learning stories for children.
- Choose this when: a small story needs an “aha” moment or an unexpected turn. Modern relationships transplanted into a prehistoric tribe make for jokes that feel both familiar and surprising.
| 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 |
A4: Cozy Storytelling
A mug too special to use. A message you feel you have to answer immediately. An afternoon finally kept for yourself. Simple characters inhabit colorful everyday settings, telling relatable stories through movement, expressions, and camera work.
- Visual style: solid white round heads, slim bodies, fine dark limbs, and natural expressions in fully illustrated 2D environments. Wide shots establish the setting, close-ups capture hesitation or relief, and tracking shots follow the character forward.
- Best for: relatable everyday stories, personal growth, relationships, heartwarming moments, habit-building, and English-learning content for children.
- Choose this when: you want viewers to recognize themselves in a small moment. Adapt the colors and lighting to the theme, from warmth and comfort to anxiety, exhaustion, and starting over.
| 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 |
A5: Historical Cartoon
Send stick figures into history: samurai on Edo streets, black cats in ancient Egyptian temples, soldiers in Roman camps. Period clothing, colorful settings, and playful contrasts make history easier to picture.
- Visual style: solid white round heads, black oval eyes, bold clear outlines, and thin line-drawn limbs, paired with period clothing, soft cel shading, and colorful backgrounds. Change the costumes, buildings, props, and music for each place and era while keeping the same character drawing style.
- Best for: world history, curious facts about civilizations, everyday life in the past, historical figures, and humorous history shorts.
- Choose this when: you want to turn a fact into something a character actually experiences. Reactions and story twists help a small historical detail hold the viewer's attention.
| 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.
B · Autopilot Mode style gallery
Styles are numbered B1, B2, B3… Each has its own finished examples combining illustrated scenes with subtle animation, maintained separately from the A series.
B1: Minimal Hand-Drawn
Let the narration lead. Expressive stick figures, colorful hand-drawn props, and bold on-screen text make ideas, stories, and product benefits clear and engaging.
- Visual style: near-white paper with a barely visible grid, black hand-drawn lines, and a few colorful props. Expressive faces, small animated details, flowing transitions, and sound effects keep the story moving. On-screen text highlights key ideas; captions follow the narration.
- Best for: YouTube explainers, psychology and personal growth, workplace stories, educational videos, and product promotion.
- Choose this when: you want a complete narrated, captioned video from a topic, copy, or story script. The visuals follow the voiceover's natural rhythm, making it a good fit for sustained explanations and long-form storytelling.
Click a GIF to view or download the preview with sound.
| Example 1: YouTube without showing your face | Example 2: Why is nobody watching? |
|---|---|
|
 |
| Hand-drawn characters · Key text · Subtle animation | Bodi voiceover · Expressive faces · Subtle animation |
B2: Coming soon…
A style overview, dedicated examples, and usage prompts will be added here.
Frequently asked questions
Q1. Can I make a video without a script?
Yes. Give the AI assistant a topic, an article, a link, or a brief, and it will shape that into a video script. You can also bring a finished script. Include your target audience, publishing platform, and intended length for a more focused result.
Q2. Can Director Mode styles be fully automated? Why keep Director Mode?
The A series currently has its own production workflows, but you can ask your AI assistant to combine your preferred A-series style with the B-series workflow for a DIY adaptation.
Director Mode is retained primarily to keep video generation and experimentation costs under control. For a video over ten minutes long, one complete generation pass can cost tens of US dollars, and repeated attempts add to the bill. That's why Director Mode keeps generation and review scene by scene, followed by manual editing: you stay in control of every shot, check the results, and then decide whether to continue.
Q3. How do I keep characters consistent in Director Mode?
Ask your AI assistant: “First design a stick figure that fits my chosen style. Let me approve the front, side, and back views before writing the storyboard prompts.” This also works if you want a character of your own. Reuse approved character references when you have them; skip this optional step if you do not need a custom design.
If your video tool supports reference images, attach the approved character image to every clip you generate. References supplement the existing storyboard prompts; they do not replace those prompts or change the selected style's rules. Check the finished clips for consistency.
Q4. Do I have to use Codex?
No. Director Mode works in AI assistants that support installing and using Skills.
Autopilot Mode needs an environment such as Codex or Claude Code that can read and write files and execute commands, plus the required image generation, voiceover services, and local dependencies.
Q5. How do I turn an article or reference into a more engaging video?
Tell your AI assistant who you want to reach, what matters most, and what you want viewers to do afterward. It can develop the angle, opening hook, narrative progression, and call to action, then turn those into scenes.
You can also share a reference video and say what you like about its pacing, delivery, or story structure. Learn from the approach while keeping your own perspective and wording. Verify historical, business, and scientific claims before publishing.
Q6. Does it cost anything?
The Skill is free and open source. Image generation, voiceover, and video generation services charge according to the platform and account plan you choose. A consumer subscription does not necessarily include API credits.
Q7. Do I have to redo everything if a visual or voiceover is wrong?
No. Approve the script and production plan first, then check a short preview for style, characters, voice, captions, and pacing before making the full video.
If only one section needs work, revise that section and reuse the approved voiceover and assets to avoid unnecessary rework and generation costs.
Q8. Can I publish on YouTube or use the videos commercially?
The project uses the MIT license, which permits commercial use. Third-party images, music, sound effects, and voiceover services remain subject to their own license terms. YouTube monetization is subject to the platform's review and policies.
More styles. More ways to create.
⭐ Building a faceless YouTube channel and exploring content monetization? Star this project for new stickman styles, video prompts, and automated workflows—so figuring out how to make the video stops holding your channel back.
🔔 Follow @CrazyKaomei for faceless YouTube workflows and practical tools that connect scripting, voiceover, image generation, and editing. Spend less time on repetitive tasks and make consistent production more manageable as a solo creator, with more ideas on topics, channel growth, and monetization along the way.
🎨 Planned styles: illustrated storybooks, doodle fights, two-character sitcoms, card-based adventures, live-action/animation hybrids, and classroom explainers.
💬 Ideas and submissions: suggest a style, share your work, or report a problem by messaging @CrazyKaomei on Telegram. For style requests, include a reference and the topic you want to cover. For submissions, include your video, the style used, and your creator link so your work can be showcased and credited.
License
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.
🇺🇸 EN · 🇨🇳 简中 · 🇯🇵 日本語 · 🇰🇷 한국어 · 🇫🇷 FR · 🇧🇷 PT-BR · 🇹🇷 TR · 🇮🇩 ID
AI 火柴人视频生成器
将任何想法变成一部高质量的、稳定画风火柴人视频。
给一个话题、一篇文章或链接，自动制作画面、旁白、字幕、音效与微动画，交付完整火柴人视频。不出镜，也能做出有辨识度的 YouTube 频道。
汇总全网热门的火柴人画风，持续更新与开源——适用于 知识科普、心理励志、教育宣传、情感治愈、流行故事等多赛道。
|
经典黑白风
|
现代潮酷风
|
|
史前洞穴风
|
治愈故事风
|
|
历史卡通风
|
极简手绘风
|
两种视频制作模式
想精细控制镜头，选 A · 导演模式；想做长片、不想逐段拼装，选 B · 全自动模式。 同一个项目，两种制作方式。
A · 导演模式 |
B · 全自动模式 |
|
|---|---|---|
| 特点 | 你来掌镜，AI 做导演：生成分镜与提示词，由你控制片段生成和剪辑。 | 给出话题或需求，直接拿成片：自动完成插画、配音、字幕、微动画与视频合成。 |
| 时长 | 短片优先，如 30 秒–3 分钟；也可逐段制作更长内容。 | 长片优先，尤其适合 10 分钟及以上的旁白视频，也能做短片和连续故事短剧。 |
| 工具 | 用支持 Skill 的 AI 助手做提案，再到 Google Flow 等平台生成片段，用剪映等工具剪辑。 | 在 Codex、Claude Code 等能读写文件、调用工具和执行命令的 AI 编程代理中运行，串起素材生成与本地合成。 |
| 画面 | 逐段生成动画：侧重角色动作、镜头调度与人物互动。 | 插画分镜＋微动画：用切镜、推拉和平移配合旁白讲故事，不必逐段生成完整动作视频。 |
| 产出 | 导演提案、分镜、逐段视频提示词与声音设计。 | 带配音、字幕与微动画的完整视频文件。 |
| 开始 |
两种方式都可从话题、文案或完整脚本开始。A、B 画风独立编号、分别展示，不共用画面；新画风与样片在各自系列下更新。 画幅、旁白语言和音色按需选择。上面的时长是选型建议，不是硬性限制。
全自动模式需要可用的图像、配音等生成服务与本地视频合成环境；普通聊天窗口不能直接替代这套运行环境。Skill 免费开源，所接服务可能另行收费。
开始使用
把下面这段发给支持 Skill 的 AI 助手。安装成功后，AI 助手会介绍两种模式能做什么、如何开始，并给出可以直接照着说的示例。
帮我安装这个 Skill：
https://github.com/kaomei/stickman-video-director
安装成功后，立即读取已安装的 SKILL.md，
用易懂的方式介绍两种模式的产出、所需工具和制作流程，
给我可以照着说的示例；
结合我的需求，告诉我还缺什么、下一步具体怎么做。
A · 导演模式：逐段生成分镜提示词
安装后，把下面这段发给 AI 助手，换成自己的话题即可：
使用导演模式/A模式，把下面的话题、文案或需求，
做成火柴人视频：
[填写话题、粘贴文案，也可以直接描述需求]
需求：横屏 16:9，采用 A3「史前洞穴风」，先做 30 秒。
先给我导演提案，确认后再输出逐段视频提示词。
流程：给出话题和需求 → 生成故事脚本 → 按需确认人物三视图 → 生成详细分镜提示词 → 手动生成片段并剪辑成片
提案包含故事结构、分段旁白、画面动作、镜头和声音设计。拿到提示词后，将它们逐条复制到 Google Flow 等视频生成工具，再用剪映等工具添加字幕、完成剪辑。
第一次做，可以跟着完整制作教程跑一遍。
B · 全自动模式：一句话生成长片
在 Codex、Claude Code 等具备文件与命令执行能力的环境中，安装后这样说：
使用全自动模式/B模式，把下面的话题、文章、帖子、新闻或具体需求，
做成在 YouTube 上有吸引力的视频：
[粘贴内容或链接，也可以直接描述宣传需求]
需求：5～10 分钟，横屏 16:9，采用 B1「极简手绘风」，配上旁白、字幕和音效。
先检查生图、配音和视频合成配置，再给我脚本与 20 秒有声样片；
确认后，让画面跟随配音节奏自动完成整片，交付完整视频文件，不用我逐段拼装。
流程：给出话题、内容或需求 → 确认视频脚本 → 确认 20 秒有声样片 → 自动制作整片 → 拿到完整视频
AI 助手会根据内容设计分镜，生成画面与旁白，配好字幕和音效，再按配音节奏合成为完整视频。你主要负责确认脚本和样片，不用逐段生成、手动拼装。
首次使用会先检查所需配置，缺什么、怎么补都会逐步说明。画面或声音不满意，可以只调整对应片段，保留已经确认的部分。
A · 导演模式画风一览
按 A1、A2、A3…更新，每个画风下展示对应的动画样片。
A1：经典黑白风
用最少的线条，把一个概念讲清楚。纯白或纯黑的画布上，人物、道具和强调色承担主要信息，适合做节奏明快的知识解说。
- 画面特点：简洁圆头、线条肢体与统一线宽，可偶尔加入简单表情；最多三种高饱和强调色，用来区分情绪、风险或行动。抽象内容可以变成具体画面，比如把焦虑画成牢笼，把注意力画成进度条。
- 适合内容：科学常识、心理成长、励志、自我管理、商业拆解与概念解释。
- 怎么选：白底黑火柴人清爽直观，适合教学和日常科普；黑底白火柴人对比更强，适合悬念、冲突和情绪表达。
| A1.1 · 白底黑火柴人 | A1.2 · 黑底白火柴人 |
|---|---|
|
|
| 纯白画布 · 黑色人物 | 纯黑画布 · 白色人物 |
A2：现代潮酷风
红色冷帽、黄色 T 恤，是这套画风最鲜明的角色标记。既可以在干净的科技空间里讲产品，也可以走进卧室、街道和荒野，演一段有起伏的故事。
- 画面特点：固定红帽黄衣的极简角色，搭配两种环境。极简科技方案用纯白空间、极淡网格、悬浮玻璃界面和青蓝光效表现科技感；彩色故事方案用彩色场景、光影和空间纵深营造故事氛围。
- 适合内容：极简科技适合 AI 工具、科技产品、方法教学与功能解说；彩色故事适合成长经历、生活话题、冒险与励志故事。
- 怎么选：重点是“讲清一个工具或方法”，选极简科技；重点是“跟着角色经历一件事”，选彩色故事。
| A2.1 · 极简科技 | A2.2 · 彩色故事 |
|---|---|
|
|
| 纯白空间 · 极淡网格 · 青蓝微光 | 彩色场景 · 电影光影 · 叙事氛围 |
A3：史前洞穴风
把今天的烦恼，搬回没有手机和办公室的史前世界。洞穴、兽皮、篝火和夸张表情，让职场、人性与生活里的荒诞，变成一段带反转的小故事。
- 画面特点：粗手绘线条、白色实心头部、深色肢体、兽皮服饰和土色环境，强调角色互动、表情反应与喜剧停顿。角色按故事设计，男女老幼都可以出现，不同故事不必共用一张脸。
- 适合内容：史前脑洞、职场讽刺、人性观察、黑色幽默、寓言与儿童英语故事。
- 怎么选：想用一个小情节讲出“原来如此”或“怎么会这样”的反转，可以选这一风格；尤其适合把现代关系放进原始部落，做出熟悉又意外的笑点。
| 案例 1：《人类的第一份工作是怎么来的？》 | 案例 2：《人类第一次发现火源》 |
|---|---|
|
|
| 史前故事 · 职场反转 | 火源主题 · 粉丝作品 |
A4：治愈故事风
一只舍不得用的杯子，一条忍不住秒回的消息，一个终于留给自己的下午。让极简角色走进彩色的日常场景，用动作、表情和镜头讲普通人能共鸣的小事。
- 画面特点：白色实心圆头与细长躯干、深色细肢体和自然的小表情，搭配完整的 2D 卡通环境。全景交代地点，近景捕捉犹豫或释然，跟拍带着角色继续向前。
- 适合内容：生活共鸣、心理成长、人际关系、温暖故事、习惯养成与儿童英语。
- 怎么选：想让观众从一个具体生活片段里认出自己，可以选这一风格。色彩和光影随主题变化，既能表达温暖，也能讲焦虑、疲惫和重新出发。
| 案例 1：通勤路上的消息焦虑 | 案例 2：把时间留给自己 |
|---|---|
|
|
| 地铁站 · 行走跟拍 · 情绪近景 | 露天咖啡座 · 暖光树影 · 日常动作 |
A5：历史卡通风
让火柴人走进历史：江户街头的武士、古埃及神庙里的黑猫、罗马营地中的士兵。用鲜明的时代服饰、彩色场景和带反差的小情节，把历史知识讲得更有画面感。
- 画面特点：白色实心圆头、黑色椭圆眼睛、清晰粗轮廓与细线肢体，搭配时代服饰、柔和赛璐璐明暗和彩色绘景。换国家、换时代时，调整服装、建筑、道具和音乐，保留同一套角色画法。
- 适合内容：世界历史、文明趣事、古人日常、人物故事与历史题材的幽默短片。
- 怎么选：想把知识点变成“古人正在经历的一件事”，可以选这一风格。既能讲一个历史细节，也能用角色反应和情节反转，让观众愿意继续看。
| 案例 1：江户日本 · 武士的日常 | 案例 2：古埃及 · 猫与木乃伊 |
|---|---|
|
|
| 江户街巷 · 武士服饰 · 日常反差 | 彩色神庙 · 黑猫角色 · 历史趣事 |
画风与提示词贡献：Lawal Ridwan / @leksautomate。
B · 全自动模式画风一览
按 B1、B2、B3…更新，每个画风下展示对应的插画分镜＋微动画成片，与 A 系列分开维护。
B1：极简手绘风
让画面跟着旁白推进。用表情生动的火柴人、彩色手绘道具和醒目的重点文字，把知识、故事或产品亮点讲得清楚又有趣。
- 画面特点：近白纸底与极淡网格，黑色手绘线条搭配少量彩色道具；用表情、局部微动画、连贯转场和音效推动画面，重点文字突出信息，字幕配合旁白。
- 适合内容：YouTube 知识讲解、心理成长、职场故事、教育科普与产品宣传。
- 怎么选：想把一个话题、一段文案或故事脚本做成带配音、字幕的完整视频，可以选这一风格。画面跟随语音节奏编排，适合持续讲解和长篇叙事。
点击动图，查看或下载有声样片。
| 案例 1：不露脸做 YouTube | 案例 2：为什么没人看？ |
|---|---|
|
|
| 手绘角色 · 重点文字 · 微动画 | Bodi 旁白 · 生动表情 · 微动画 |
B2：待更新…
画风介绍、独立样片与使用示例将在这里更新。
常见问题
Q1. 没有脚本，也能做视频吗？
可以。给一个话题、一篇文章、一个链接或具体需求，AI 助手会先整理成视频脚本；已有脚本也可以直接使用。最好一起说明目标观众、发布平台和期望时长。
Q2. 导演模式的画风可以做成全自动吗？为什么要有导演模式？
目前 A 系列画风的制作流程相对独立，但你可以让 AI 助手采用喜欢的 A 系列画风，结合 B 系列的制作流程，自行 DIY。
保留导演模式，主要是为了控制视频生成和试错成本——制作十几分钟的长视频，一轮完整生成的费用可能达到 几十美元左右，反复重做还会增加开销。因此，导演模式保留逐段生成、检查和手动剪辑，既让你掌控每个镜头，也能先确认效果，再决定是否继续。
Q3. 导演模式生成的人物不一致怎么办？
可以先告诉 AI 助手：“先帮我设计符合所选画风的火柴人，确认三视图（正面、侧面、背面），再给分镜提示词。”想做专属形象的视频，也可以这样开始。已有确认的角色参考图可直接复用，不需要定制时也可跳过这一步。
在支持参考图的视频工具中，生成每段视频时都附上已确认的角色图。参考图是额外辅助，不替代原有分镜提示词，也不改变所选画风的规则；最终仍需检查各段人物是否一致。
Q4. 一定要用 Codex 吗？
不一定。导演模式可以在支持安装和使用 Skill 的 AI 助手中使用。
全自动模式需要 Codex、Claude Code 等能够读写文件、执行命令的环境，并配置相应的生图、配音服务和本地依赖。
Q5. 怎么把文章或参考内容，变成更有吸引力的视频？
告诉 AI 助手你想吸引谁、重点讲什么，以及希望观众看完后做什么。AI 助手会据此提炼选题角度，设计开场钩子、叙事推进和结尾引导，再整理成分镜。
有喜欢的视频，也可以提供参考，说明你喜欢它的节奏、表达方式或故事结构。借鉴方法，保留自己的观点与表达；涉及历史、商业和科学的事实，发布前仍需核查。
Q6. 使用需要付费吗？
Skill 免费开源。生图、配音和视频生成等服务，按你选择的平台及账号方案计费；订阅某个平台的会员，不一定包含它的 API 调用额度。
Q7. 画面或配音不满意，需要全部重做吗？
不需要。建议先确认脚本和制作方案，再做短样片，检查画风、人物、声音、字幕与节奏，满意后再制作整片。
如果只有某一段不合适，可以优先调整对应片段，复用已经确认的配音和素材，减少返工与重复消耗。
Q8. 做出来的视频可以发 YouTube、用于商业项目吗？
项目采用 MIT 许可，允许商用。视频使用的第三方图片、音乐、音效及配音服务仍需遵守各自的授权条款；能否开通 YouTube 收益，则以平台审核为准。
更多画风，更多创作可能
⭐ 想做不露脸的 YouTube 频道，探索内容变现？ Star 这个项目，持续获取火柴人画风、视频提示词和自动化制作流程，让你的频道不再卡在“视频怎么做”这一步。
🔔 关注 @CrazyKaomei：获取不出镜 YouTube 视频的自动化工作流与实用工具，串联脚本、配音、生图和剪辑，减少重复操作，让一个人也能持续产出。更多选题、频道增长与内容变现思路，持续分享。
🎨 更多画风（计划中）：手绘绘本、涂鸦打斗、双人情景喜剧、卡牌闯关，以及虚实融合和讲台解说。
💬 建议与投稿：想看新画风、分享作品或反馈问题，欢迎通过 Telegram 联系 @CrazyKaomei。画风建议请附参考和想做的主题；投稿请附作品、所用风格与作者链接，方便展示和署名。
许可证
本项目采用 MIT License 开源，允许免费使用、修改、商用和分发。分发本项目或其重要部分时，请保留原有版权和许可声明。
项目按“现状”提供，不作任何担保；完整条款见 LICENSE。