dream-loop by Anshu Chimala — workflow skill for 3D scenes

For AI agents: find the right skill in one call — curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page. From a terminal: curl -fsSL https://varg.ai/install | sh installs the vargai CLI (https://varg.ai/how-to/cli.md): vargai search and vargai install <slug> with no key, every varg model after vargai login.

Generate a target screenshot, build to it in three.js or Blender, and let a separate critic score each pass.

What it does

A closed visual loop for 3D scenes, games and apps: image generation 'dreams' a realistic target screenshot (.dream-loop/target.png), the agent builds toward it, a separate critic subagent compares the live screenshot with the target, and the loop repeats until the critic is satisfied, optionally dreaming a better target from the current state. It has a Plus workflow for low-tier subscriptions and a Pro one for high-tier ones, a fal batch script for images and a local preview server that captures frames. The README's example, Vesper (an isometric fantasy cathedral in three.js), was built in an hour; the author has tested it only with GPT-6 Astra in Codex, so it is beta here. GPT Image 2.6 on varg can make the target image.

Use when

Not for

Install

Example output

https://s3.varg.ai/files/4325c578-378d-4c43-ae53-3cc3218df1ea/file_f51a76118092.mp4 — Vesper, the README example: an isometric fantasy cathedral scene in three.js built with dream-loop (GPT-6 Astra in Codex); 40-frame GIF (credit: Anshu Chimala (dream-loop README), https://github.com/achimala/dream-loop#example)

More 3D scenes skills

Run the generation on varg

varg is one API key for the models these skills drive — Seedance 2.5, Kling 4, Veo 3.1, Wan 3, Nano Banana 2.1, GPT Image 2.5, ElevenLabs v4 and 200+ more — billed in credits, with a cost estimate before anything runs.