Docker Sandboxes – Disposable, isolated sandboxes for AI agents

Need to enforce these controls across your whole team?

Network and filesystem controls you define.

Enforceable org-wide with Docker AI Governance.

Hard security boundary from the host.

Disposable by default. Faster than VMs.

Agents can spin up containers within Sandboxes.

Install packages, run services, work unattended.

Claude Code, Gemini CLI, Copilot CLI, Codex, Kiro, OpenCode.

A sandbox is a microVM isolated environment that protects your filesystem and network from agents running inside it.

Out of the box we support Claude Code, Gemini CLI, Copilot CLI, Codex, OpenCode, Kiro. You can also create your own

YOLO mode (–dangerously-skip-permissions) gives agents autonomy with no approval prompts. Essential for speed, but risky without guardrails. Sandboxes make it safe by isolating each agent inside a dedicated microVM.

Sandboxes run fully isolated in microVMs, giving more isolation without paying the full cost of running a VM. This lets them do things that need more permissions safely, like running additional Docker containers.