Scenario 3d Worlds

Organization

scenario-labs

scenario-3d-worlds

Use when generating navigable 3D worlds or Gaussian splat scenes on Scenario via MCP: text-to-world, image-to-world, a 360 panorama or skybox turned into a walkable splat, multi-view photos or a video walkthrough reconstructed as a splat, or one object photo to splat. Keywords: World Labs Marble 1.1, 1.1 Plus, 1.0 Draft, Hunyuan World, HY World, image-to-skybox, skybox-to-splat, multiview-to-splat, TripoSplat, gaussian splatting, 3DGS, spz, world generation, explorable scene, VR.

Overview

Publisher	scenario-labs
Repository	`skills`
Skill name	`scenario-3d-worlds`
Stars	970
Forks	114
Bundled files	Instructions only
License	MIT
Links	Skill source Repository Homepage

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.
Self-contained
Everything the model needs lives in the instructions — no extra files to sync.
Open source
Published by scenario-labs on GitHub. Read the source before you install it.

Installation

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

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.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/scenario-labs/skills/tree/main/skills/scenario-3d-worlds
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.

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/scenario-labs/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/scenario-3d-worlds .claude/skills/scenario-3d-worlds

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Scenario 3d Worlds 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 Scenario 3d Worlds 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 Scenario 3d Worlds 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.

Scenario 3D Worlds

Overview

These models build explorable environments as 3D Gaussian splats, not polygon meshes: volumetric scenes to walk through and render in real time, with no UVs or topology to edit (Marble splats also export to triangle or collider mesh). Three vendors sit side by side: World Labs Marble in three quality tiers, Hunyuan's HY World pipeline, and TripoSplat for single objects. Discover them with search and treat model_schema_get as the contract; HY World members carried a geo restriction tag at authoring time, so hits differ by team and region.

Connection and the core loop: see the scenario skill; polygon meshes, rigging, and the 3D viewer: the scenario-3d skill; skybox-only work with no splat stage: the scenario-skyboxes skill. If a sibling skill named here is missing from your available skills, ask the user to install it (npx skills add scenario-labs/skills --skill <name>); unattended, proceed from tool schemas and flag the gap.

Quick reference

Pick the member by what you already have (input names from the live schemas):

You have	Member	Inputs
A text idea	Marble, any tier	`prompt` alone
One image, flat or 360	Marble	`images`, one entry; `isPano: true` if 360
2 to 8 photos of one scene	Marble	`images`
A walkthrough video	Marble or HY Multi-view to Splat	`video`
Many overlapping photos	HY Multi-view to Splat	`images`, 2 to 64 at authoring time
One photo of a place	HY Image to Skybox, then Skybox to Splat	`image`, then `panorama`
One photo of an object	TripoSplat	`image`; background removal is automatic

images and video are mutually exclusive wherever both exist; isPano applies only when exactly one image is passed. TripoSplat wants one clean object and every other member wants a scene: aiming either at the other's subject is the main failure mode.

Draft on Marble, then upgrade

The three Marble tiers (1.0 Draft, 1.1, 1.1 Plus) take identical inputs, and the same inputs with the same seed reproduce the same world on every tier. Iterate on Draft, which returned in about a minute at authoring time, then rerun the identical parameters on 1.1 (balanced) or 1.1 Plus (highest fidelity, dynamically grows larger worlds when the scene allows). The cost spread across tiers was more than tenfold at authoring time, so dry_run before committing. splatResolution picks the stored density (100k, 500k, full); disableRecaption: true stops Marble rewriting your prompt.

The HY World pipeline runs long

Image to Skybox is an image model, not a 3D one: it expands one photo of a place into a 2:1 equirectangular panorama (backend trades full fidelity against qwen speed and cost). Its output asset id goes to Skybox to Splat as panorama. There, the three trajectory toggles (applyNavTraj, applyUpRoute, applyReconIteration, default true) drive cost and time; maxSteps sharpens marginally and materially moves neither. The splat stage ran close to an hour at median at authoring time: keep re-calling jobs_wait with pending_job_ids; a timeout is not a failure and never justifies a second model_run. Multi-view to Splat is the opposite, fast and cheap: best when the camera moved through a static scene with parallax.

Worked example: graybox a level from a concept image

search with target="models", query="marble world", public=true. Tiers appear side by side, e.g. model_worldlabs-marble-1-0-draft (a live hit at authoring time: re-discover each session).
model_schema_get with the Draft id: fields, caps, and defaults before anything else.
upload_asset the concept image (see the scenario skill) for an asset id.
model_run with dry_run=true and parameters={"images": ["asset_x"], "prompt": "abandoned observatory interior, dusk light through the dome slit", "seed": 1234, "splatResolution": "full"}; repeat the dry run on the 1.1 id to see the upgrade cost.
model_run on Draft with wait=false, then jobs_wait with the job id, re-called with pending_job_ids on timeout.
asset_display the splat and walk it; adjust and repeat on Draft until the layout is right.
Rerun the identical parameters, same seed, on the 1.1 or 1.1 Plus id from step 1, then asset_download the final .spz.

Common mistakes

Passing images and video together: mutually exclusive on Marble and Multi-view to Splat.
A single equirectangular input without isPano: true: Marble reads it as a flat photo.
Changing or omitting seed between the draft and the upgrade tier: a different world comes back.
Feeding TripoSplat a scene, or a world model a lone object.
Treating a jobs_wait timeout on Skybox to Splat as a failure: the job is still running.
Adjusting maxSteps to buy quality or cut cost: the trajectory toggles move cost, maxSteps barely moves sharpness.
Expecting UVs or editable topology from a splat: for polygon assets, use the mesh generators in scenario-3d.

Frequently asked questions

What does the Scenario 3d Worlds AI skill do?

Why use Scenario 3d Worlds on TypingMind?

Because you install it once and use it with any model. Scenario 3d Worlds 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 Scenario 3d Worlds in TypingMind?

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

Which AI models can use Scenario 3d Worlds?

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 Scenario 3d Worlds?

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

Is the Scenario 3d Worlds AI skill free?

Yes. It is published on GitHub by scenario-labs 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

Nemo Curator

NousResearch

OrganizationPopular

Curate LLM training data: dedupe, filter, PII redaction.

Obliteratus

OBLITERATUS: abliterate LLM refusals (diff-in-means).

Prompt Lookup

Activates when the user asks about AI prompts, needs prompt templates, wants to search for prompts, or mentions prompts.chat. Use for discovering, retrieving, and improving prompts.

Banner Design

Design banners for social media, ads, website heroes, creative assets, and print. Multiple art direction options with optional generated or supplied visuals. Actions: design, create, generate banner.…

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Scenario 3D Worlds

Overview

Quick reference

Draft on Marble, then upgrade

The HY World pipeline runs long

Worked example: graybox a level from a concept image

Common mistakes

Frequently asked questions

Related AI Skills

Set up your own AI workspace now