Minimalist Monochrome	Modern Tech
Prehistoric Dark Comedy	Cozy Storytelling
Historical Cartoon	More styles coming soon…

Find your style

Style 1: Minimalist Monochrome

A few lines can explain a big idea. On a pure-white or pure-black canvas, characters, props, and accent colors do the storytelling. A good fit for fast, clear explainers.

• The look: Round heads, line limbs, consistent stroke weight, and occasional simple expressions. Up to three vivid accent colors distinguish emotions, risks, or actions. Turn anxiety into a cage or attention into a progress bar.
• Best for: Science, psychology, motivation, self-management, business breakdowns, and explaining concepts.
• Which version? White backgrounds feel clean and direct; choose black backgrounds for stronger contrast, tension, and emotional impact.

Style 1A: Black on White	Style 1B: White on Black
White canvas · Black figure	Black canvas · White figure

---

Style 2: Modern Pop

The red beanie and yellow T-shirt make this character easy to recognize. Use a clean tech setting to explain a product, or take the character into bedrooms, streets, and open landscapes for a story.

• The look: One recognizable character, two kinds of world. 2A pairs white space and a faint grid with floating glass interfaces and cyan-blue light. 2B uses colorful settings, cinematic lighting, and depth.
• Best for: 2A suits AI tools, tech products, tutorials, and feature demos. 2B suits personal growth, everyday situations, adventures, and motivational stories.
• Which version? Choose 2A to explain a tool or method; choose 2B to follow a character through an experience.

Style 2A: Minimal Tech	Style 2B: Cinematic Story
White space · Faint grid · Cyan glow	Colorful settings · Cinematic light · Story atmosphere

---

Style 3: Prehistoric Sketch / Dark Comedy

Take today's problems back to a world without phones or offices. Caves, animal hides, campfires, and exaggerated expressions turn familiar workplace frustrations and human quirks into stories with a twist.

• The look: Rough hand-drawn outlines, white-filled heads, dark stick limbs, animal-hide clothes, and earthy scenery. Character interaction, reaction shots, and comic timing matter. Cast adults, children, or elders to fit the story; every story needn't use the same face.
• Best for: Prehistoric what-ifs, workplace satire, observations about human nature, dark comedy, fables, and children's English stories.
• Why choose it? For a short setup and an unexpected payoff—especially when modern relationships become absurdly familiar in a prehistoric tribe.

Example 1: “How Did Humans Get Their First Job?”	Example 2: “Humans Discover Fire”
Prehistoric workplace comedy	Fire discovery · Community creation

---

Style 4: Cozy Storytelling

A mug you've been saving for a special occasion. A message you feel compelled to answer. An afternoon you finally keep for yourself. Simple characters step into colorful everyday settings, with actions, expressions, and camera work carrying the story.

• The look: White-filled round heads, slender bodies, dark stick limbs, and small, natural expressions in full-color 2D cartoon environments. Wide shots establish the place; close-ups catch hesitation or relief; tracking shots keep the story moving.
• Best for: Relatable everyday stories, personal growth, relationships, warm moments, habit building, and children's English content.
• Why choose it? To help viewers recognize themselves in a small moment. Color and lighting follow the topic: warmth, anxiety, exhaustion, or a fresh start.

Example 1: Message Anxiety on the Commute	Example 2: Making Time for Yourself
Train station · Tracking shots · Emotional close-ups	Outdoor café · Warm light · Everyday gestures

---

Style 5: Historical Cartoon

History, with a cast you can follow. Send a samurai through Edo's streets, meet a black cat in an Egyptian temple, or follow a soldier into a Roman camp. Period costumes, colorful settings, and small comic turns make the past feel like a story rather than a slide deck.

• The look: White-filled round heads, simple black oval eyes, bold outlines, and thin stick limbs, with period clothing, soft cel shading, and painted backgrounds. Change the wardrobe, buildings, props, and score for each era—not the underlying character style.
• Best for: World history, surprising facts, everyday life in the past, historical figures, and history-themed comedy.
• Why choose it? To turn a fact into something happening to a character. Use a clear situation, a telling reaction, or a twist to carry the explanation.

Example 1: Everyday Life in Edo Japan	Example 2: Cats and Mummies in Ancient Egypt
Edo streets · Samurai clothing · Everyday surprises	Painted temple · Black cat · Historical curiosity

Style and prompt contribution: Lawal Ridwan / @leksautomate. Japanese sample provided by the original creator.

What the Skill does

It plans the writing and the visuals together—not just a stick figure's appearance.

Part	What you get
Script	A hook, progression, and payoff, with narration and a reference translation where needed.
Visuals	Character actions, props, camera moves, transitions, and sound cues.
Continuity	Character, prop, and narrator profiles, plus transitions between clips.
Framing	Composition and shot choices adapted to landscape or portrait.
Revisions	A proposal to approve before the final prompts.

First, review the director's proposal: narration, scenes, camera work, and sound design. Once you approve it, get standalone Gemini Omni Flash video prompts for each segment, plus stitching, transition, and subtitle notes.

The Skill plans the video and writes the prompts. Flow or another video tool generates the clips; a video editor handles the final assembly and subtitles. No API or MCP setup is required for the Skill.

Get started

1. Install the Skill

Send this to your AI assistant:

Install this Skill for me:
https://github.com/kaomei/stickman-video-director

2. Give it a topic

Send a topic, draft, or article. For example:

Make a stickman video called “How Did Humans Get Their First Job?”

3. Follow the prompts and approve the plan

Choose settings → Review the proposal → Request changes → Approve and get your prompts.

If you haven't chosen an aspect ratio or style, the Skill asks. Pick landscape 16:9 or portrait 9:16, then a style from the gallery. It won't ask again for settings you've already given. Videos default to 60 seconds; say “let's try 30 seconds first” for a shorter test.

Next comes the director's proposal: story structure, narration for each segment, character actions, camera work, and sound design. Review the whole plan and request changes before approving the final video prompts.

“How Did Humans Get Their First Job?” | Landscape 16:9 · Style 3: Prehistoric Dark Comedy · 30 seconds

This is the storyboard approved before making the example above. The sample's original narration is preserved:

Time	Scene and action	Voiceover
0–10s: Keeping trouble out	A caveman drags his stone club into the cave and collapses onto a straw bed. A huge saber-toothed tiger's shadow appears at the entrance. He opens one eye: “Again?” He rolls a boulder across the opening, dusts off his hands, and goes back to bed. End on his satisfied face.	“Ten thousand years ago, one man finally solved all his problems. He shut them outside.”
10–20s: The boss finds you	Just as he closes his eyes, three knocks come from the entrance. The boulder shifts open a crack. The chief leans in, smiling, and offers a spear. The caveman waves it away. Then the chief produces a roasted leg of meat. The caveman sniffs, eyes suddenly wide. Cut outside: spear in hand, he's being nudged toward giant animal tracks.	“Then someone offered him food… in exchange for a small favor. Congratulations. Employment was invented.”
20–30s: The price of progress	He returns to the same cave covered in dust, drops the spear, and holds out his hand for the meat. The chief offers a tiny piece with a professional smile, then presses a whole bundle of spears into his arms, pats his shoulder, and leaves. The caveman looks from the meat to the spears, then slowly at the camera. He lies back down, but the spears jab him upright.	“The reward got smaller. The workload got bigger. And somehow… they called that progress.”

Don't like a line or a scene? Say what to change. Final prompts come only after your approval. This 30-second example produces three roughly ten-second prompts covering characters, action, camera work, narration, and audio.

4. Generate the clips and edit your video

Submit the prompts to Google Flow in order. Assemble the clips in CapCut or another editor, adjust the joins, and add subtitles. A 30-second video usually uses three roughly ten-second prompts; a 60-second video uses six.

For a first run, follow the complete production walkthrough. There's also a Flow tutorial. Both are in Chinese. The Skill is free; video generation costs and model availability depend on the platform.

A few practical tips

Landscape or portrait?

Format	Good for	Composition
9:16	YouTube Shorts, TikTok, Instagram Reels	Vertical action, a central subject, and mobile-safe framing
16:9	YouTube and landscape educational videos	Lateral movement, wider scenes, and side-by-side comparisons

Use it for faceless explainers, history and business topics, personal growth, everyday stories, and product demos. English narration with bilingual subtitles added later also works for English-learning content, short dramas, and close-reading videos. Durations can be any multiple of ten seconds: 30 seconds, 60 seconds, 90 seconds, three minutes, and beyond.

• Keep character, prop, and narrator profiles consistent across the whole video. Choose any voice changes in the proposal first.
• If one clip's character, props, or voice drift, adjust and retest that clip rather than remaking everything.
• Improve the writing without inventing facts, statistics, or product claims.

Latest updates

• 2026-10-20: Made English the default homepage and added Turkish and Indonesian pages. Chinese, Japanese, Korean, and Brazilian Portuguese remain available. Skill rules unchanged.
• 2026-10-20: Added Style 5, Historical Cartoon, with Japanese and Egyptian samples and the Japanese film in the hero preview. Thanks to @leksautomate for the style and prompt contribution.
• 2026-10-15: Refined the Chinese homepage, installation steps, and examples; added lighter previews and a Star History chart. Skill rules unchanged.
• 2026-10-14: Added Style 4, Cozy Storytelling, with train-station and café examples.
• 2026-10-13: Refined Style 1's character descriptions and current-pose constraints.
• 2026-10-12: Improved shared voice-consistency rules and added prehistoric and community examples.

Already installed? Homepage changes alone don't require a reinstall. Update the Skill if you want Style 5 or installed before the character and voice-rule changes. Detailed changelog (Chinese).

What should we add next?

These styles are planned, not available yet:

• Mixed-media worlds — Exploring 2D stick figures in richer environments.
• Blackboard Explainer — Lessons and talks with blackboard interaction.
• Style 6: Illustrated Storybook — Fairy tales, fables, and children's stories.
• Style 7: Graffiti Action — Fight scenes, short action films, and combos.
• Style 8: Two-Character Sitcom — Banter, everyday misunderstandings, and workplace twists.
• Style 9: Card-Game Adventure — Level-ups, monster battles, and game-like storytelling.

Suggest a style, share your work, or report a problem in Issues. For style requests, include a reference and a topic you'd like to try. For submissions, include the video, style used, and an author link so we can credit you.

Star History

⭐ If this library saved you time hunting for prompts or fixing failed generations, consider giving it a Star. It'll be here when you're ready to make your next video.

I'm Kaomei. I'll keep testing styles and sharing useful prompts, workflows, and the mistakes worth avoiding—all for free. Less time wrestling with tools, more time for the stories you actually want to tell.

License

Released under the MIT License. You can use, modify, redistribute, and use the project commercially. Keep the copyright and license notice when distributing the project or substantial portions of it.

Provided “as is,” without warranty. See LICENSE for the full terms.