TL;DR
A new open-source engine allows teams to deploy self-hosted, isolated development sandboxes with live preview URLs using Docker and Go, eliminating the need for Kubernetes. It supports multiple users, cost-efficient operation, and easy setup, making it suitable for scalable AI app builders and coding platforms.
An open-source engine named sandboxed now provides a self-hosted solution for creating isolated development environments with live preview URLs, using only Docker and Go, without requiring Kubernetes.
Sandboxed is designed to run on a single server, creating private Linux containers that are spun up via a simple API call. These containers include AI coding agents and live dev servers, accessible through shareable URLs, with automatic handling of TLS and routing via Traefik. The platform is built with a minimal stack: a Go program orchestrates Docker containers, SQLite manages state, and Traefik handles URL routing, making it straightforward to understand and deploy.
It supports use cases like multi-user AI app builders, coding playgrounds, per-user or per-branch preview environments, and team-based multi-app hosting. The system stops idle sandboxes to conserve resources and wakes them on demand, enabling cost-effective operation. The installation process is simplified to a single command, ./install.sh, after which the API becomes accessible at localhost:90. The platform includes pre-installed AI agents such as OpenCode and Claude Code, capable of generating code within sandboxes based on prompts.
Why It Matters
This development matters because it offers a scalable, cost-efficient, and easy-to-deploy solution for teams needing multiple isolated development environments. By avoiding Kubernetes, it reduces complexity and resource requirements, making it accessible for smaller teams or those wanting to maintain full control over their infrastructure. It addresses common challenges like environment isolation, URL routing, and resource management, which are critical for AI app builders, coding platforms, and collaborative development workflows.
Podman in Action: Secure, rootless containers for Kubernetes, microservices, and more
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Traditional multi-tenant development environments often rely on complex Kubernetes clusters or multiple virtual machines, which can be costly and difficult to manage. The recent rise of AI app builders and coding playgrounds has increased demand for lightweight, self-hosted solutions that can handle multiple users efficiently. Prior efforts have focused on cloud-hosted solutions or manual container management; sandboxed offers a streamlined, self-contained alternative that leverages Docker’s simplicity, combined with a minimal control plane built in Go.
“Sandboxed is designed to be the simplest way to run many isolated dev environments on a single server, with minimal setup and maximum scalability.”
— Sandboxes developer team
“This tool could significantly lower the barrier for teams wanting to run their own AI app builders without investing in complex infrastructure.”
— Hacker News user ‘techenthusiast’
self-hosted development environment server
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
Details about long-term stability, security, and performance in production environments remain unconfirmed. It is also unclear how well the platform scales beyond a single server or how it integrates with existing CI/CD pipelines. The extent of support for different cloud providers or environments is still to be tested.
live preview URL hosting solutions
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Next steps include broader testing in production environments, community feedback, and potential feature additions such as multi-host support or enhanced security features. Developers may also explore integrating the platform with existing development workflows or expanding AI agent capabilities.
Codex Remote: Engineering From Your Phone: The Definitive Guide to OpenAI's Coding Agent — CLI, Cloud, MCP, and the Mobile Control Plane
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Can I run sandboxed on my own server?
Yes, sandboxed is designed for self-hosting on a single server with Docker installed. The setup process involves cloning the repository and running a simple installation script.
Does it require Kubernetes or other orchestration tools?
No, sandboxed operates solely with Docker and does not depend on Kubernetes, making it simpler to deploy and manage.
What are the main use cases for this platform?
It is suitable for running multiple AI app builders, coding playgrounds, per-user or per-branch preview environments, and team-based multi-application hosting.
Is this platform suitable for small projects or just large-scale deployments?
Sandboxed is overkill for small, one-off projects. It is best suited for teams or platforms managing many sandboxes simultaneously.
Source: Hacker News
