YGYOOO/WorldX

One sentence creates an AI-driven world — generate maps, characters, and watch stories emerge on their own. 一句话生成一个AI自主驱动的世界.

What it solves

WorldX allows users to create a fully realized virtual world from a single sentence. It eliminates the need for manual world-building, character scripting, or map design by using AI to generate the environment, inhabitants, and the emergent stories that unfold between them.

How it works

The project uses a multi-model pipeline to transform a text prompt into a simulation:

  1. Orchestration: An LLM designs the world structure, rules, and characters.
  2. Generation: Image models create the map art and character portraits, while a vision model reviews the map quality and identifies elements.
  3. Simulation: A simulation engine (powered by LLMs) drives autonomous agents who make decisions, interact with the environment, form relationships, and maintain memories.
  4. Frontend: A Phaser 3 and React 19 client renders the world as a pixel-art simulation where users can observe or intervene as a "God" figure.

Who it’s for

It is designed for creators, simulation enthusiasts, and developers interested in emergent narratives and autonomous AI agents who want to quickly prototype complex virtual societies.

Highlights

  • One-sentence creation: Generates entire maps and character sets based on a simple description.
  • Autonomous Agents: Characters possess memory and personality, allowing them to evolve and interact without pre-written scripts.
  • God Mode: Users can inject events, edit character memories/personalities, or engage in direct dialogue with agents.
  • Temporal Evolution: Supports day-night cycles and multiple diverging timelines for the same world.
  • Multi-Model Architecture: Separates roles into Orchestrator, Image Gen, Vision, and Simulation models for optimized performance.