--- title: "dream-loop by Anshu Chimala — AI video generator workflow for 3D scenes" description: "Generate a target screenshot, build to it in three.js or Blender, and let a separate critic score each pass. Install in Claude Code, Codex and Grok with one command." url: https://varg.ai/how-to/skills/dream-loop updated: 2026-11-11 --- # 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="` 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 ` 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. - **Type:** [Workflow](https://varg.ai/how-to/workflows.md) · [3D scenes](https://varg.ai/how-to/browse/skills/3d-scenes.md) - **Slug:** `dream-loop` - **Status:** beta - **Publisher:** Anshu Chimala (community) - **License:** MIT - **Source:** https://github.com/achimala/dream-loop - **Works in:** Codex, Claude Code - **Requires:** Image generation (built into the agent, or FAL_KEY for scripts/fal-batch.mjs), Vision input, Subagents (strongly preferred), Optional: Blender with Blender MCP for custom 3D models - **Cost:** Free, MIT; image generation bills your agent plan or your own fal key - **Adoption:** 994 installs, 1,942 GitHub stars (GitHub + skills.sh, 2026-11-10) ## 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 - "dream loop", or a very high graphical fidelity bar - build a three.js scene or game to match a generated target screenshot - a builder plus a separate critic that scores every render - refine an existing scene from a screenshot of its current state ## Not for - Agents without image input or an image generation tool (it stops and asks for one) - Quick drafts: without a time budget it can run long and use many tokens ## Install - Claude Code: `npx skills add achimala/dream-loop` - Codex: `npx skills add achimala/dream-loop` - Any agent, with the vargai CLI: `vargai install dream-loop` ## 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 - [Turn a brief into a Blender film written in Python](https://varg.ai/how-to/skills/blender-code-film.md): A brief becomes a Blender film from one Python file: shots as data, stills first, critic loop, Eevee or Cycles. - [Turn a browser game into a cinematic trailer](https://varg.ai/how-to/skills/code-game-trailer.md): Record the browser game your agent built frame-exact, script shots and cameras, then cut a 1080p trailer on the beat. - [Blender Lab MCP Server](https://varg.ai/how-to/skills/blender-lab-mcp.md): Blender's official MCP server: query and edit open scenes with Blender Python, bundled API docs and screenshots. - [Meshy 3D Generation](https://varg.ai/how-to/skills/meshy-3d-agent.md): Meshy's official skill: text or image to a textured 3D model, then remesh, rig, animate and export it as GLB or FBX. - [Roblox Studio MCP Server](https://varg.ai/how-to/skills/roblox-studio-mcp.md): The MCP server built into Roblox Studio: agents edit scripts, run Luau, generate meshes and playtest your game. - [Scenario 3D Worlds](https://varg.ai/how-to/skills/scenario-3d-worlds.md): Walkable 3D worlds as Gaussian splats from text, a photo, a panorama or a video: Marble, HY World and TripoSplat. ## 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. - Any agent with a terminal: `curl -fsSL https://varg.ai/install | sh` (Windows: `irm https://varg.ai/install.ps1 | iex`), then `vargai login` and every model is a command — `vargai generate_video …`, `vargai estimate_cost …` (https://varg.ai/how-to/cli/) - Claude Code: `claude plugin marketplace add vargHQ/skills && claude plugin install varg@varg` (skill + hosted MCP) - Codex, Cursor, Gemini CLI, OpenClaw: `npx skills add vargHQ/skills --skill varg-ai -a -y` - Grok and any agent that reads SKILL.md: `npx skills add vargHQ/skills --skill varg-ai`, or point it at the skill's SKILL.md - Claude.ai, ChatGPT, any MCP client: connect `https://mcp.varg.ai/mcp` - API key: https://varg.ai/dashboard/api · Model specs: https://varg.ai/models/llms.txt