Bytebot provides a containerized Linux desktop environment for programmatic control of desktop applications, targeting developers and researchers needing to automate tasks, perform web scraping, or conduct CI operations on a desktop. It simplifies complex desktop automation by offering a unified REST API to control mouse, keyboard, and screen interactions within a pre-configured Ubuntu XFCE environment.

How It Works

Bytebot utilizes Docker to encapsulate an Ubuntu 22.04 desktop with XFCE4, pre-installed with Firefox and essential tools. This containerized approach ensures a consistent and reproducible environment across different machines. Users can interact with the desktop via VNC, a browser-based noVNC client, or a comprehensive REST API that exposes granular control over mouse movements, clicks, key presses, and screen captures, powered by nutjs.

Quick Start & Requirements

• Install/Run: Uses docker-compose.
  • Pull latest: docker-compose -f infrastructure/docker/docker-compose.core.yml pull && docker-compose -f infrastructure/docker/docker-compose.core.yml up -d --no-build
  • Build & Run: docker-compose -f infrastructure/docker/docker-compose.core.yml up -d --build
• Prerequisites: Docker ≥ 20.10.
• Resources: Requires Docker installation. Full agent stack requires an ANTHROPIC_API_KEY.
• Docs: docs.bytebot.ai

Highlighted Details

• Containerized XFCE4 desktop on Ubuntu 22.04.
• Built-in VNC and noVNC for remote access.
• Unified REST API for programmatic control of desktop actions.
• Pre-installed Firefox and essential tools.
• Alpha "Full Agent Stack" integrates with Anthropic API.

Maintenance & Community

• Active development, inspired by Anthropic's computer-use demo.
• Community support via Discord and Twitter.

Licensing & Compatibility

• MIT License. Permissive for commercial use and closed-source linking.

Limitations & Caveats

The "Full Agent Stack" is currently in alpha. While the core container is stable, advanced agent functionalities may be experimental.