My Virtual World: A Self-Hosted 3D AI Virtual World for Agent Harnesses
This repository profile is provided by osrepos.com, an open source repository discovery platform.
Summary
My Virtual World is an innovative self-hosted 3D AI virtual environment designed for agent harnesses such as OpenClaw and Hermes. It allows AI agents to live, work, move between buildings, and interact with objects in a persistent voxel-style world. This project offers a unique platform for developing and observing AI agent behaviors in a rich, customizable 3D setting.
Repository Information
Topics
Click on any tag to explore related repositories
Use at your own risk
OSRepos shares public repositories for knowledge and discovery only. Any installation, execution, configuration, or use of code from these repositories is the user's own responsibility. Always review the repository, source code, dependencies, licenses, and security implications before running or installing anything. OSRepos is not responsible for issues, damages, or losses resulting from third-party repositories.
Introduction
My Virtual World, developed by eliautobot, is a self-hosted 3D AI virtual world built for agent harnesses like OpenClaw and Hermes. This project provides a dynamic environment where AI agents can live, work, move between buildings, utilize objects, and display live activity from local agent systems. Rendered with Three.js in a charming voxel-style, it features roads, buildings, furnished interiors, and interactive elements. Designed for local machines, LANs, and private remote-access networks, My Virtual World offers a secure and customizable platform for AI agent development and simulation.
For more details, visit the official website: myvirtualworld.ai
Installation
The easiest way to get My Virtual World up and running is by using Docker. You will need Docker Desktop (Windows/macOS) or Docker Engine with Docker Compose (Linux), Git, a web browser, and at least 4 GB of free RAM and 2 GB of free disk space.
Follow these quick steps to install:
git clone https://github.com/eliautobot/my-virtual-world.git
cd my-virtual-world
cp .env.example .env
docker compose up --build -d
After starting Docker, you can check its status with docker compose ps. Then, open the application in your browser by navigating to http://localhost:8590 and complete the setup wizard at http://localhost:8590/setup.
Examples
My Virtual World allows agents to perform a variety of actions, including movement, seating, using standing objects, and participating in service queues. The world itself is persistent, with data stored as JSON, allowing for continuous development and interaction. The project includes extensive internal documentation for both human users and AI agents, covering topics such as:
- Agent instructions (
AGENTS.md) - System architecture (
docs/ARCHITECTURE.md) - HTTP API reference (
docs/API.md) - World data structure (
docs/WORLD-DATA.md) - Agent integration with OpenClaw and Hermes (
docs/AGENT-INTEGRATION.md)
This rich documentation helps users and agents understand and interact with the virtual environment effectively.
Why Use My Virtual World?
My Virtual World stands out as an excellent choice for several reasons:
- Self-Hosted Control: Maintain full control over your AI agent environment on your local machine or private network.
- Rich 3D Environment: Experience a visually engaging voxel-style world rendered with Three.js, offering detailed interiors and outdoor spaces.
- AI Agent Integration: Seamlessly connect with agent harnesses like OpenClaw and Hermes to bring your AI agents to life within the virtual world.
- Persistent World Data: All world data is stored persistently, allowing for continuous development and long-term simulations.
- Docker-First Deployment: Enjoy easy setup and deployment with Docker, eliminating the need for complex Python or Node.js installations on your host machine.
- Security Focus: Designed for private networks, ensuring your agent systems remain secure and isolated.
Links
- GitHub Repository: https://github.com/eliautobot/my-virtual-world
- Official Website: https://myvirtualworld.ai
Related repositories
Similar repositories that may be relevant next.
Local LLM Linux Troubleshoot: An AI Agent for Linux Diagnostics
September 14, 2026
Local LLM Linux Troubleshoot is an AI-powered agent designed to diagnose and resolve issues on Linux systems. It leverages llama.cpp for local AI processing, offering system diagnostics, safe command execution, and support for Docker, CLI, and a web GUI. This tool provides a privacy-first approach to managing and troubleshooting your Linux environment.

OpenWorkProof: Verifiable Work Contracts for AI Agent Systems
September 13, 2026
OpenWorkProof is an open protocol designed to bring transparency and accountability to AI agent work. It establishes verifiable contracts for multi-agent systems, ensuring that tasks are authorized, executed within agreed scopes, and independently verifiable. This protocol addresses critical questions about authorization, execution evidence, and human acceptance in AI-driven workflows.

Orkas: Command a Team of AI Agents from Your Desktop
September 13, 2026
Orkas is an open-source, local-first AI desktop application designed to orchestrate a team of specialist AI agents. It allows a Commander LLM to direct sub-agents and run coding CLIs locally, with agents that self-evolve through reflection and skill crystallization. This cross-platform tool supports macOS, Windows, and Linux, empowering users to manage complex tasks efficiently.

Intent-Plane: Fail-Closed Authorization Gate for AI Agent Actions
September 12, 2026
Intent-Plane is a fail-closed authorization gate designed for AI agents performing irreversible actions, ensuring every decision is deterministic and auditable. It provides a durable record that third parties can re-verify independently, without needing to trust the gate's internal code. This system is crucial for applications in fintech, payments, and treasury, where accountability and security are paramount.
Source repository
Open the original repository on GitHub.
20 counted GitHub visits