image-blaster is a Claude skillset that generates 3D environments, sound effects, and meshes from a single image using World Labs and FAL APIs.
The tool solves the problem of rapid 3D asset creation for game development, architectural visualization, and other 3D workflows. It works by accepting an image as input and using Claude as an orchestrator to coordinate multiple generation models. The process separates dynamic objects into 3D models in GLB or OBJ format, renders the static environment as a Gaussian splat, and generates ambient and object-specific sound effects. The entire pipeline can complete in under five minutes, significantly accelerating the initial stages of 3D project setup.
The tool suits developers and creators who need to quickly prototype 3D environments or jumpstart asset production without extensive manual modeling work. It integrates with major game engines including Unity, Unreal, and Godot, as well as digital content creation software like Blender, 3DS Max, and Maya. Web-based projects using Three.js or Electron can also embed the generated assets. The approach offers granular control over 3D model generation through parameters like target face count, PBR material generation, polygon reduction, and polygon type selection. Use cases range from video game level concepts and architectural renderings to film location scouting and environment design for robotics applications.
The project shows active development with recent commits and ongoing refinement of its generation pipeline. The codebase is written in TypeScript and includes a React viewer component that developers can modify by adjusting configuration files. The tool maintains integration with multiple external generation models, suggesting ongoing maintenance of API compatibility and feature parity across different generation backends.
