Skip to content

neilsonnn /image-blaster

Turn one image into a 3D set you can reuse across every shot
Agent skillsMITEnglish README
Stars
13k
30 days
+1.1k
Last push
6 months ago
image-blaster screenshot from its README
From the image-blaster README on GitHub.

About image-blaster

A set of Claude skills that turns one image into a 3D scene. You get a walkable 3D environment, 3D models of the objects in it, and matching sound effects, in under five minutes. For an AI drama, that means one location you can film from any angle, so the room stops changing between shots. The catch: you need Claude Code plus paid World Labs and FAL keys, and the repo hasn't been updated since May 2026.

Best for: Creators who keep fighting a set that looks different in every shot, and who already use Claude Code.

See it in action

Turn a Single Image into a Fully Meshed 3D World by World Labs, on YouTube. It's two minutes of the real run: one photo in, a 3D room, objects and sound out. It's not ours.

What it does

  • Builds a walkable 3D environment (a Gaussian splat, .spz) from one image
  • Makes 3D models (.glb, .obj) of the moving objects in the scene
  • Adds a looping ambient sound and per-object physics sound effects (.mp3)
  • Asks you to confirm each step when you say blast it and confirm each step with me
  • Exports into Unity, Unreal, Godot, Blender, Maya or a Three.js app
  • Lets you set the 3D model detail, from 40,000 to 1,500,000 faces

Quickstart

Rewritten from the README. Check the repo for the latest steps.

  1. 1
    Clone the repo: the command below
    git clone https://github.com/neilsonnn/image-blaster
  2. 2
    Go in: cd image-blaster
  3. 3
    Start Claude Code: claude
  4. 4
    Give Claude your World Labs and FAL API keys when it asks
  5. 5
    Put an image in input/ and say: the command below
    blast it and confirm each step with me

Models and languages

Models it works with

  • World Labs Marble
  • Nano Banana
  • GPT Image 2
  • Hunyuan 3D
  • ElevenLabs

Interface and docs

  • English

Alternatives

Other agent skills projects people compare with image-blaster.

NomaDamas

CozyClay

3D blocking and camera moves, exported as keyframe packs for AI video

Storyboard
  • Seedance 2.5
  • Kling 2
  • Veo 3
  • +6
1.8k GitHub stars+1k/30dtodayAGPL-3.0

wassermanproductions

motion-previs-studio

Feed it a reference shot, get pose, depth and camera-move files

Storyboard
  • Seedance
  • Runway
  • Kling
386 GitHub stars+20/30d4 months agoApache-2.0

eternityspring

shuohao-skills

Pre-production assets for an AI short drama, straight from a novel

Agent skills
  • Qwen Image
  • GPT Image 2
  • MiniMax H3
  • +1
6.7k GitHub stars+2k/30d8 days agoApache-2.0

zenstory-ai

drama-skills

Chains eleven agent skills into your whole AI short-drama pipeline

Agent skills
  • MiniMax H3
  • Seedance
  • GPT Image 2
  • +1
4.4k GitHub stars+1.4k/30dyesterdayMIT

Featured on OpenMicroDrama

Do you maintain image-blaster? Add this badge to your README so English speakers can find our write-up.

Featured on OpenMicroDrama
markdown
[![Featured on OpenMicroDrama](https://openmicrodrama.com/badges/featured.svg)](https://openmicrodrama.com/projects/neilsonnn-image-blaster)

The badge links to this page. Want your description changed? Email support@openmicrodrama.com.

Reviewed Oct 4, 2026. Added 2026-10-04 on request. 7.2k stars, MIT, last push 2026-05-15. Demo video is World Labs' own walkthrough; fal's 12-minute tutorial is jGwnOGLzMl0. We write these descriptions ourselves. The repo's own docs are the final word.