add-3d-models

Add 3D GLTF models to Decentraland scenes with bounding-box validation and collider setup.

3|Updated Apr 13, 2026
One-click install
npx skills add https://github.com/decentraland/sdk-skills --skill add-3d-models-decentraland
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: add-3d-models
Source: https://github.com/decentraland/sdk-skills/tree/main/add-3d-models
Command: npx skills add https://github.com/decentraland/sdk-skills --skill add-3d-models-decentraland

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Adding 3D GLB/GLTF models to a Decentraland scene is error‑prone because models often have hidden offsets, unexpected scales, and missing colliders, leading to invisible or misplaced assets.

Core Features & Use Cases

  • Bounding‑box validation: script to compute true world‑space size accounting for node transforms.
  • Collider configuration: guidance on visible vs invisible mesh collision masks and _collider detection.
  • Composite integration: add static models to main.composite and dynamic models via GltfContainer.
  • Placement safety: rules for safe positioning, orientation of single‑sided models, and text label visibility.
  • Model sourcing: workflow for searching the OpenDCL catalog, downloading assets, and referencing them in code.

Quick Start

Use the add-3d-models skill to place a tree model at position (2, 0, 2) with appropriate colliders.

Frequently Asked Questions about add-3d-models

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add 3D GLTF models to a Decentraland scene without them being misplaced?

To add 3D GLTF models to a Decentraland scene safely, you need bounding-box validation to compute true world-space size and correct placement calculations. This prevents hidden offsets and unexpected scales from causing invisible or misplaced assets.

How do I configure colliders for GLB models in Decentraland?

Collider configuration for GLB models in Decentraland involves setting up visible versus invisible mesh collision masks and detecting `_collider` nodes. Proper collider setup ensures physical interactions match the visual boundaries of the imported 3D assets.

What is the difference between static composites and dynamic runtime entities for 3D models in Decentraland?

Static composites integrate 3D models into `main.composite` for fixed placement, while dynamic runtime entities use `GltfContainer` for flexible loading. Both require proper bounding-box validation and collider configuration to function correctly in the scene.

Can I use models from the OpenDCL catalog in my Decentraland scene?

Yes, you can use models from the OpenDCL catalog in your Decentraland scene by searching the catalog, downloading the assets, and referencing them in your code. The workflow includes proper bounding-box validation and composite file edits for accurate placement.

Why does my 3D model have invisible or misplaced assets when imported into Decentraland?

Invisible or misplaced 3D assets in Decentraland typically occur because models have hidden offsets, unexpected scales, and missing colliders. Bounding-box validation scripts and placement safety rules are required to resolve these import errors.

Do I need a script to measure the size of a GLTF model in Decentraland?

Yes, an optional script for size measurement is often needed to compute the true world-space size of a GLTF model in Decentraland. This bounding-box validation accounts for node transforms to ensure accurate placement calculations.