HELLO, CREATORSGood work starts with a great skill.Find yours

Stickman Video Director

kaomei/stickman-video-director

Turn copy, notes, or a topic into a stickman director’s proposal, English narration, and standalone video prompts. Choose a visual style, 16:9 or 9:16, and a duration in ten-second increments.

Latest update at:

2,143Stars316ForksRepository last pushed

About this Skill

Stickman Video Director installs the directing-stickman-videos skill and its reference files into your agent. Start with copy, notes, an article, or a topic. The agent plans narration, physical actions, camera changes, transitions, music, and sound effects, then waits for approval before producing the final prompts. A sixty-second plan normally contains six approximately ten-second clips. The deliverable is a proposal and prompt package; you generate and assemble the video in your chosen tools.

Why we picked it

Useful for creators who want to plan an explainer or story before generating clips. The skill separates proposal review from final prompt delivery and carries character, prop, and voice descriptions across clips. This listing links to kaomei’s original MIT repository; both animations are the author’s demonstrations.

What it helps with

  • Explain science, AI tools, product ideas, or mental models through concrete visual actions.
  • Plan motivational stories, personal-growth shorts, workplace satire, or gentle everyday narratives.
  • Recompose a script for vertical Shorts/Reels/TikTok or a horizontal explainer.
  • Review story and voiceover choices before spending credits on video generation.

What the skill produces

Phase A is a director’s proposal in your conversation language, with English narration and a reference translation. It lays out the message, hook, tone, visual beats, camera, transitions, music, and sound effects. Phase B supplies standalone prompts after you approve the current proposal.

The duration defaults to sixty seconds. Choose a multiple of ten seconds, such as thirty, ninety, or one hundred eighty seconds. Each prompt describes an approximately ten-second clip, including timed actions, exact narration, continuity anchors, and an ending that connects to the next segment.

Installable Skill entrypointDirector proposal templateVideo prompt contract

Choose a visual style

The current skill entrypoint and style catalog include the following modes. The two previews above show Studio Tech and Prehistoric Hand-Drawn. The English README’s older roadmap differs from the current skill files; the descriptions here follow the installable entrypoint and references.

Classic Minimalist

A black figure on white or a white figure on black, with a small set of accent colors for props and graphics.

Modern Beanie: Studio Tech / Cinematic Story

A red-beanie, yellow-shirt character in a white studio with cyan UI, or in full-color story environments.

Prehistoric Hand-Drawn / Dark Comedy

Earthy scenery, animal-hide costumes, expressive casts, and fictional visual analogies with setup, escalation, and an ironic payoff.

Cozy Storytelling

Minimalist expressive characters in full-color 2D cartoon locations, with planned location changes and character/prop continuity.

Installable Skill entrypointCurrent style catalogPrehistoric storytelling referenceCozy Storytelling reference

Plan your first video

  1. Provide the source and settings

    Paste your source and specify 16:9 or 9:16, a visual style, and optionally duration and voice. Missing aspect ratio or style triggers a setup question; omitted duration defaults to sixty seconds.

  2. Review the proposal

    Check the narration, scene actions, shot changes, and audio plan. For example: explain perfectionism in thirty seconds, 16:9, Studio Tech, with a calm narrator.

  3. Approve the current version

    Ask for revisions while the plan is still text. A change to ratio, duration, style, narration, or global direction returns the workflow to proposal review.

  4. Generate clips and assemble

    Submit approved prompts to your video tool, review the results, and stitch the clips. Add subtitles in post-production; default prompts request no visible writing.

Installable Skill entrypointDirector proposal template

Character, prop, and voice continuity

The skill repeats character descriptions and a complete approved voice profile in each prompt. It tracks prop ownership, opening and ending states, screen direction, and planned cuts. Vertical and horizontal versions receive different staging rather than a ratio-label swap.

Repeated instructions guide generation but do not guarantee identical anatomy, faces, music, or voices. Review actual footage and listen across joins. One continuous external voiceover and music track can help when independently generated audio varies.

Installable Skill entrypointVoice continuity referencePrehistoric storytelling referenceCozy Storytelling reference

Install the original skill

Use the command in the installation panel and select your agent host. Keep the full directing-stickman-videos folder, including references. The repository’s own instructions also describe cloning it and copying skills/directing-stickman-videos into Codex’s skills directory, then restarting Codex and invoking $directing-stickman-videos.

Official repository and installationInstallable Skill entrypoint

Generation and evidence

This is a planning and prompt-writing skill. It does not bundle a video renderer or automate Flow, editing, or publication. The previews are compressed copies of the author’s supplied GIF demonstrations, shown in the requested order. This listing review covers the source instructions and marketplace presentation; it does not certify newly generated footage or every compatible agent.

Official repository and installationAuthor’s Studio Tech animationAuthor’s prehistoric animationMIT license

Before you use it

  • An agent that can load SKILL.md and its local references. The repository documents Codex installation; the shared skills installer can target other compatible hosts.
  • No API key or MCP server is required for the proposal and prompt-writing workflow. The npx installation command requires Node.js and network access.
  • Use a separate video-generation service for the clips and an editor to assemble them. The author’s prompts target Gemini Omni Flash; check your provider’s current model availability and pricing.
  • The skill is free under MIT. Agent/model usage and video generation may carry separate charges.

Supported agents

Claude CodeCodex CLICursorGitHub CopilotDeepSeek HarnessGemini CLIGrok BuildHermes AgentOpenCodePiKiroRovo DevQoderMistral VibeGoogle Antigravity

Frequently asked questions

Does it export a finished MP4?

It delivers the approved proposal and prompt package. Generate clips in a separate service, then assemble them in your editor.

Can I write the brief in Chinese?

Yes. Phase A follows the user’s conversation language, with English voiceover and a reference translation. Final generation prompts are written for the author’s English narration workflow.

Is the skill free?

The repository is MIT-licensed and free to install. Your agent and chosen video-generation provider set their own usage charges.

Sources & references

Original sources are linked below. Practical starting points are editorial suggestions from SkillHey.