Animated Explainer

Big canvas mind map

A big hand-drawn mind map with the camera following each branch

Picture shape: 9:16
Runs on Claude CodeCodexAntigravity
Watch on YouTube

Explain one topic made of many connected parts (an ecosystem, a product family, a tech stack, a business system) in a 90-180 second vertical video (1080x1920), drawn as one big mind map.

Important: this is a skill for terminal AI coding agents (Claude Code, Codex) that runs on your own computer. It is not a prompt for the ChatGPT website. Check the requirements below before buying.

Why this look

Some topics are not one idea but a set of parts that connect. Here there is one map, one pen and one camera: a hand draws the hub, then each branch in its own colour while the camera travels from node to node. Nothing is cut, and the last move pulls back to show the whole map.

  • Each node is inked on the word the voice says.
  • Lists, small trees, cycles, cards and red "the point" notes are written under each node.
  • Calm AI voice, lo-fi acoustic music with no drop.

The story of the video

  1. Hook: a question on a sticky note; the hub is the only thing on the map.
  2. 6 to 10 branches: one part each. Name it, say what it is for in one picture, end on the point.
  3. Recap: the camera pulls back to the whole map, and a yellow sticky lists the parts with a one-line takeaway.

Simple parts come first, the part that ties them together last. You choose the number and order of branches; the layout is automatic.

How it works

Open your agent and type a request like this:

Make a mind map video of the parts of an online shop and how they connect

The agent writes the facts (names must be verified) and the script in its own words, anchors every reveal on a spoken word, generates the voice-over in one take, fits the music, then renders with HyperFrames and checks the file: 1080x1920, 30 fps, -14 LUFS. Two review gates before render need approval from another model, a subagent or a person.

What you get

  • The skill folder, with project templates, the storyboard format and the render and check scripts.
  • Per video: the MP4, a cover PNG and a copy of the video with the cover on its first frames.

What you need

  • Claude Code or Codex.
  • Node.js 20 or newer, ffmpeg and ffprobe, ImageMagick 7, Python 3.10 or newer with numpy and scipy, and multix (npm install -g @mrgoonie/multix).
  • ELEVENLABS_API_KEY for music and sound effects; GEMINI_API_KEY for the default English voice. SONIOX_API_KEY is optional.
  • Vietnamese voice: free and local, installed once with python scripts/setup-vieneu.py (about 1.1 GB, runs on CPU).

Install

Copy the skill folder into:

~/.claude/skills/agentvid-motion-big-canvas-mindmap-9x16/

For Codex, use ~/.agents/skills/agentvid-motion-big-canvas-mindmap-9x16/.

Who it is for

Explaining one topic made of many connected parts as one pen-drawn map.

Not a fit if you want

  • To explain a single idea, or a process step by step.
  • To edit real footage or make a sourced documentary.

Frequently asked questions

Which languages does it support?

Vietnamese and English. The voice and every word on the map follow the language of your input unless you ask for another.

Can I use my own brand?

Yes. Name, logo and colours go in one brand.json; logos must be transparent PNGs.

Can I use the videos commercially?

Videos you render are yours for any commercial use. You may not resell, redistribute or share the skill files.

Reviews

What buyers say

No reviews yet

No reviews yet.

More

Related looks