Visualize In 3d logo

Visualize In 3d

Organization
codepawl
visualize-in-3d

Turn anything hard to picture into a 3D world people can watch or play with - an operation or system (airport, data center, port, factory, hospital), an event spreading over real land (wildfire, flood), a story in time (a historical event, a case), characters or machines in action (a fight, a chase, a launch), or how a thing works. Builds a live three.js app with a dashboard and a 30 s film, a lesson app, a quick no-Blender prototype, a live embed that runs on real data inside an existing dashboard, a documentary film, an action short, or an interactive cutaway plate or book. Looks include toy isometric, point cloud, photoreal terrain, dark documentary, pixel diorama and illustrated cutaway plate. Models come from ready sources or Blender builds made from real references, and the skill checks its own frames before showing anyone. Use for "visualize how X works", "show this event on a 3D map", "a live 3D dashboard of X", "a 3D lesson on how X works", "a documentary-style 3D explainer", "a quick 3D prototype", "two creatures fight in 3D", "app nhìn như game", "làm phim 3D về", "cho 2 con đánh nhau", "pixel battle", "theme point cloud", "kiểu wildfire tool".

Overview

Publishercodepawl
Repositoryviz3d
Skill namevisualize-in-3d
Stars
89
Forks
8
Bundled files
24
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 24 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by codepawl on GitHub. Read the source before you install it.

Installation

Install the Visualize In 3d AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/codepawl/viz3d/tree/main/plugins/visualize-in-3d/skills/visualize-in-3d
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/codepawl/viz3d.git /tmp/viz3d
mkdir -p .claude/skills
cp -r /tmp/viz3d/plugins/visualize-in-3d/skills/visualize-in-3d .claude/skills/visualize-in-3d
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Visualize In 3d in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Visualize In 3d on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Visualize In 3d is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Visualize in 3D (Dựng thế giới 3D để thấy mọi thứ vận hành)

If references/local.md exists, read it first: it sets the folders and house rules for this machine.

Take any subject and make it visible in 3D: models from real references, a seeded simulation or a scripted timeline driving them, a look chosen for the subject, and a film checked frame by frame. The user can also start it with /viz3d <mode> <subject> (modes: app, lesson, prototype, embed, film, action, book, game). Paths use {{WORK_DIR}} for the folder where projects live (the current directory unless the user names one) and {{SKILL_DIR}} for this skill's folder. Write paths with forward slashes on every platform.

Before you start

  1. Blender (not needed for a quick prototype, an embed, or the pixel look): run blender --version or look in the usual install folder. If it is missing, say it gives much better models, is free at blender.org, and ask whether to wait or go on without it. Without it, follow references/no-blender.md and say in the report that model quality is lower.
  2. Node 20+ and Playwright with Chromium for recording; ffmpeg for encoding and checks.
  3. Subagents: where the harness has them, give each pass to a fresh subagent with its brief and keep it inside about 80 tool calls. Where it does not, run the passes one after another in this session, finish and check one before starting the next, and re-read the plan file at the start of each.
  4. Downloads (elevation, map data, models, reference images, sounds): name the source, count and rough size, and get the user's OK first.

1. Pick the shape

Read what the user wants people to see, then pick one shape (or combine two) and read its reference:

The subject isShapeRead
A place or system that runs (units, tasks, flows)Live app with dashboard and a 30 s filmreferences/brief.md, examples/
How a thing works (aircraft, circuit board, engine)Lesson app: 3 to 5 lessons, one control each, a "Build your own" sandboxreferences/brief.md (parts, flow overlays, sandboxes)
Something spreading over real landTerrain app with a time scrubberreferences/looks.md (photoreal terrain), examples/wildfire-valley.md
A story, case or piece of historyDocumentary filmreferences/documentary.md, references/film.md
Characters, creatures or machines doing somethingAction shortreferences/action.md
One object cut open or explodedCutaway film, or an app with a section viewreferences/film.md, references/looks.md
One big object or building with many rooms, as a page or a bookBook: interactive illustrated cutaway plate, with quizzesreferences/book.md, references/looks.md
A game people play in the browser (steer a creature, survive, grow)Playable web game, bot-balancedreferences/game.md
A demo for leaders in about an hourQuick prototype from three.js primitivesreferences/live-embed.md
A live view inside an existing dashboard or productEmbed (web component or iframe) on real datareferences/live-embed.md, scripts/mock-feed.mjs

Pick the look in references/looks.md (Choosing): clean ops world (default for a system that runs), toy isometric, point cloud, photoreal terrain, dark documentary pixel diorama or illustrated cutaway plate. Study a look the user shows from a reference video with timestamped frame sheets (scripts/contact-sheet.sh <video> <folder> <fps>), then write it up as a new section in looks.md.

2. Write the plan file

Write one file in {{WORK_DIR}} before anything is built: the topic file for an app (the examples/ shape), the fact file for a documentary, the action sheet for an action short. It names the real parts of the subject, what moves or changes, the beats or features to show, the look, and the camera view that shows the subject best (a cutaway for depth, top-down for flows, low side for arrivals).

Research the real subject first: real dimensions, steps, rules and moves. Use made-up names for companies and people unless the user asks for real ones. For a real named place, follow "A real place" in references/real-objects.md.

3. Pass 1: assets and stage

Models follow references/asset-sources.md: ready models first, then scans, packs, AI-generated models, and Blender builds from a reference sheet (references/real-objects.md). The pass reports stills at film scale. A review loop that reads many frames runs them through a frame-reader agent; open only the frames a decision hangs on.

4. Pass 2: build and film, with the self-review loop

Build the app or the film from the plan file with the brief in references/brief.md. Before reporting, run references/self-review.md for up to three rounds: scripts/frame-guard.js measurements, frame sheets and the mode's fault list. Run the two passes separately; one pass doing both runs too long.

5. Check, then show

Open the final sheets and peak stills yourself (scripts/contact-sheet.sh, sheets newer than the video). Look for what the loop can miss: a subject wrong for the real thing, a dull beat, broken UI rules, toys or programmer art. Every fault the user or you find that the loop missed goes into the fault lists the same turn (self-review.md, Growing the lists).

6. Polish

Give fixes to one more pass with "everything else is approved, change nothing else", then re-record. Publish only on the user's word.

Rules

  • Original names and designs by default. Real brands, liveries and known characters only when the user asks, with licensed models and a credit line.
  • Every main asset matches a reference of the real thing. Models from memory read as toys.
  • Every frame is a function of time (a virtual clock, clips set from it), so any frame renders the same in any order. Real clicks on that clock drive an app's film, not a camera moved in code.
  • Render and encode at low priority, one heavy job at a time, and stop every server you start.
  • Real people and real harm: facts from sources, victims named with respect, no gore (references/documentary.md).
  • Reference images and videos are for study only: never ship them in the app or the film.
  • The public copy of this skill is mirrored from this folder. Edit only this folder; never put personal paths, names or tools outside references/local.md. A nightly sync script mirrors the rest.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Visualize In 3d AI skill do?

Turn anything hard to picture into a 3D world people can watch or play with - an operation or system (airport, data center, port, factory, hospital), an event spreading over real land (wildfire, flood), a story in time (a historical event, a case), characters or machines in action (a fight, a chase, a launch), or how a thing works. Builds a live three.js app with a dashboard and a 30 s film, a lesson app, a quick no-Blender prototype, a live embed that runs on real data inside an existing dashboard, a documentary film, an action short, or an interactive cutaway plate or book. Looks include...

Why use Visualize In 3d on TypingMind?

Because you install it once and use it with any model. Visualize In 3d is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Visualize In 3d in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/codepawl/viz3d/tree/main/plugins/visualize-in-3d/skills/visualize-in-3d. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Visualize In 3d?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Visualize In 3d?

As many as you like. As long as a model supports skills, you can use Visualize In 3d with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Visualize In 3d AI skill free?

Yes. It is published on GitHub by codepawl under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇