#4 · Primary category: Agent-Based Modeling & Simulation

WorldX

agent ai ai-agents ai-characters ai-simulation ai-world emergent-behavior game-ai generative-agents generative-ai llm llm-agents multi-agent-simulation open-source phaser pixel-art typescript

One sentence creates an AI-driven world — generate maps, characters, and watch stories emerge on their own.

Project last updated:07/08/26

GitHub Stars

1.4K

Forks

233

Contributors

3

License

MIT

Why we included this project

WorldX turns a single sentence into a running virtual world. Describe a place and who lives there, and the system generates the map art, character sprites, and a simulation in which agents make decisions, remember what happened, build relationships, and talk to each other with no scripted storyline. The architecture is genuinely modular: four separate roles (orchestrator, image generation, vision review, and a runtime simulation engine) each expose their own API endpoint and model, so you can swap in the LLM or image provider you already use. It is alpha software and you will hit rough edges, but for anyone curious about emergent narrative or multi-agent behavior, the pixel-art sandbox makes the simulation visible in a way plain text logs never do, with a timeline system to branch and compare runs and a god mode to inject events or edit agent memories.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category