sandbox
All-in-One Sandbox for AI Agents that combines Browser, Shell, File, MCP and VSCode Server in a single Docker container.
The AIO Sandbox MCP server is a Docker-based unified execution environment that bundles browser automation, shell access, file operations, Jupyter notebooks, VSCode Server, and multiple pre-configured MCP servers into a single container. It provides a shared filesystem across all components, enabling AI agents to seamlessly coordinate complex workflows like downloading files via browser and immediately processing them with shell commands. The server exposes both REST APIs and native MCP protocol endpoints, making it compatible with any MCP client while also supporting direct SDK integration via Python, TypeScript, and Go libraries.
Key Features
Use Cases
- 01Web scraping workflows that combine browser automation with file processing and data analysis
- 02AI agents that need to download files via browser and immediately process them with shell or Python scripts
- 03Complex automation tasks requiring coordination between browser actions, code execution, and file manipulation
- 04Development and testing of AI agent behaviors in a fully isolated, reproducible sandbox environment
- 05Document conversion workflows combining browser rendering, screenshot capture, and markdown generation
- 06Multi-step data pipelines that require browser interaction, code execution, and file system operations in sequence
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
sandbox — FAQ
What is the AIO Sandbox MCP server?+
It's an all-in-one sandbox environment packaged as a Docker container that provides browser automation, shell access, file operations, and multiple pre-configured MCP servers through a unified API. All components share the same filesystem, enabling seamless workflows across different tools.
How do I install and run the AIO Sandbox?+
Run the Docker container with the command 'docker run --security-opt seccomp=unconfined --rm -it -p 127.0.0.1:8080:8080 ghcr.io/agent-infra/sandbox:latest'. For client SDK usage, install via 'pip install agent-sandbox' (Python), 'npm install @agent-infra/sandbox' (JavaScript), or 'go get github.com/agent-infra/sandbox-sdk-go' (Go).
Which MCP clients work with AIO Sandbox?+
The sandbox exposes MCP endpoints at http://localhost:8080/mcp and can be connected to any MCP-compatible client. It also provides native SDKs for programmatic access from Python, TypeScript, and Go applications.
Do I need API keys to use AIO Sandbox?+
No external API keys are required for basic functionality. You can optionally set SANDBOX_API_KEY environment variable to protect access to the sandbox services (API, JupyterLab, VNC). Without it, services remain open for backward compatibility.
Is the AIO Sandbox free and open source?+
Yes, AIO Sandbox is released under the Apache License 2.0 and is completely free to use. The Docker image is publicly available on GitHub Container Registry.
What are the system requirements for running AIO Sandbox?+
You need Docker installed with support for the '--security-opt seccomp=unconfined' flag. The container requires at least 2GB of shared memory (configurable via shm_size) and binds to port 8080 by default.
How do I install sandbox?+
Open the source repository on GitHub and follow its README. sandbox is a mcp server — MCP Agents Market links you directly to the official repo.
Is sandbox free?+
sandbox is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.