Docker-based Desktop Environment for GUI Agents, ScreenEnv, Released

ScreenEnv is a Python library that builds an isolated Ubuntu desktop environment inside a Docker container to test and deploy GUI agents (Computer Use agents). It provides a sandbox environment where agents can interact with real applications without complex VM setup.

Key features are as follows:

• Full desktop control: Supports mouse and keyboard automation, window management, application launching, file operations, terminal access, and screen recording.
• Two integration methods: Supports both a Direct Sandbox API for custom frameworks and a Model Context Protocol (MCP) server mode for integration with AI systems, enabling flexible architecture design.
• Docker-native: Supports AMD64 and ARM64 architectures, and can build a reproducible environment in under 10 seconds, making deployment simple.
• smolagents support: Natively integrates with Hugging Face's smolagents, allowing rapid creation of custom desktop agents.