Sandbox FAQs

Question 1

What is Sandbox and how does it differ from containers or VMs?

Accepted Answer

Sandbox is a lightweight, OS-native tool for isolating processes on macOS and Linux. Unlike containers or virtual machines, it uses kernel-level features (like `sandbox-exec` and `bubblewrap`) to provide robust security with minimal overhead, near-native performance, and instant startup, without the need for a separate guest OS or runtime.

Question 2

What platforms does Sandbox support?

Accepted Answer

Sandbox is cross-platform, providing consistent process isolation capabilities across both macOS (using `sandbox-exec`) and Linux (using `bubblewrap`).

Question 3

What kind of security does Sandbox provide for untrusted code?

Accepted Answer

Sandbox offers secure-by-default dual isolation: filesystem controls prevent unauthorized data access (e.g., SSH keys), and network controls block unrestricted outbound connections. This protects your system from untrusted code, such as AI agents or third-party scripts, by restricting its access to only what's explicitly allowed.

Question 4

Can I configure Sandbox for specific access control needs?

Accepted Answer

Yes, Sandbox is highly configurable. You can define fine-grained access policies for network connectivity (allowed/denied domains) and filesystem operations (deny read, allow/deny write paths) via a simple JSON configuration file or its Node.js library, ensuring precise control over sandboxed processes.

Question 5

What are the key benefits of using Sandbox over other isolation methods?

Accepted Answer

Sandbox offers minimal performance overhead with near-native speeds, blazingly fast startup times (milliseconds), and extremely low memory consumption (<1MB per process). It achieves strong kernel-level isolation without the complexity or resource demands of full containers or virtual machines, making it ideal for highly efficient and secure execution of untrusted code.

Sandbox

Sandbox

Key Features

Use Cases

Key Features

Use Cases