Im a graduate student doing research in the Artifical Intelligence domain at the moment. With this project, wanted to see how far I could push agentic coding assistants and I know 3d rendering is one of the hardest systemic tasks to get right.

The result is a sort of first person demo I have that supports save/load "scenes", normal mapping, mesh and texture caching, render-device abstractions and more.

I’m releasing this under the MIT License. Given the current Supreme Court discussions around AI and copyright, I believe that if the authorship of agent-assisted code is legally "thin," the only logical move is to make the architecture a public resource.