> For the complete documentation index, see [llms.txt](https://docs.ai.neevcloud.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.ai.neevcloud.com/agentic-studio/overview.md).

# Sandbox

Sandbox gives you on-demand, isolated compute environments that you can spin up in seconds. You can choose a base template, configure your resources, and get a fully provisioned environment ready to run AI agents and code - without any infrastructure setup or long-term commitment.

With Sandbox, you can quickly experiment, build, and deploy in a controlled environment while maintaining flexibility and efficiency.

Sandbox is designed for running autonomous AI agents and code execution workloads in secure, isolated environments - see [Sandbox Use Cases](/agentic-studio/overview/getting-started/use-cases.md) for exactly when to reach for it, who typically uses it, and what they build.

New here? Start with [Serverless Compute for Agents](/agentic-studio/overview/integrations-and-support/serverless-compute.md): why an agent should never run on your own machine, and how a sandbox keeps your secrets in, unknown hosts out, idle time free, and every mistake undoable.

## Sandbox Dashboard

When you open Agentic Studio from the left sidebar and navigate to Overview, you see a real-time summary of your sandbox pool:

* **Total Sandboxes** – Total number of sandboxes in the selected project, across all states
* **Ready** – Number of sandboxes that are healthy and reachable
* **Needs Attention** – Sandboxes that require action, such as failed provisioning
* **Templates** – Available base images, including standard and browser templates

## Sandbox Templates

Every sandbox is created from a template. A template defines the base operating system and pre-installed tooling your sandbox starts with. NeevCloud provides the following default templates:

Each template has an id and a name. Pass the **id** as `sandbox_template_id` when you create a sandbox — that is also the value you see back on the sandbox record.

| Template ID               | Name                   | Type     | Description                                                                  |
| ------------------------- | ---------------------- | -------- | ---------------------------------------------------------------------------- |
| `sb-ubuntu-26-04-dev`     | `ubuntu-26.04-dev`     | Standard | Ubuntu 26.04 with development tooling preinstalled. **The platform default** |
| `sb-ubuntu-26-04-minimal` | `ubuntu-26.04-minimal` | Standard | Ubuntu 26.04 minimal base                                                    |
| `sb-ubuntu-24-04-minimal` | `ubuntu-24.04-minimal` | Standard | Ubuntu 24.04 minimal base, stable and widely supported                       |
| `sb-debian-12-minimal`    | `debian-12-minimal`    | Standard | Lightweight Debian 12 base for general-purpose agent workloads               |
| `sb-ubuntu-26-04-systemd` | `ubuntu-26.04-systemd` | Standard | Ubuntu 26.04 with systemd, for workloads that require service management     |
| `sb-ubuntu-26-04-vnc`     | `ubuntu-26.04-vnc`     | Browser  | Ubuntu 26.04 with VNC, for web-based and browser agent tasks                 |

**Tip:** Omit `sandbox_template_id` entirely and you get the platform default, which is the batteries-included `sb-ubuntu-26-04-dev`. Choose a minimal base instead when you want a smaller image and are installing your own tooling anyway.

## Where to Go Next

| I want to...                                                                   | Go to                                                                                                                   |
| ------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------- |
| Understand what makes Sandbox safe for agents                                  | [Serverless Compute for Agents](/agentic-studio/overview/integrations-and-support/serverless-compute.md)                |
| See when, why, and by whom Sandbox is used, and the typical path to production | [Sandbox Use Cases](/agentic-studio/overview/getting-started/use-cases.md)                                              |
| Create my first sandbox                                                        | [Getting Started](/agentic-studio/overview/getting-started/getting-started.md)                                          |
| Connect to a running sandbox from my own code                                  | [Connecting to a Sandbox](/agentic-studio/overview/development-and-connectivity/connecting-to-a-sandbox.md)             |
| Read, write, or manage files inside a sandbox                                  | [Working with Files](/agentic-studio/overview/development-and-connectivity/working-with-files.md)                       |
| Run a command or start a long-running process                                  | [Running Commands & Processes](/agentic-studio/overview/development-and-connectivity/running-commands-and-processes.md) |
| Share a live preview of something my agent built                               | [Exposing Ports for Preview URLs](/agentic-studio/overview/development-and-connectivity/exposing-ports.md)              |
| Control which hosts a sandbox can reach                                        | [Internet Access and Egress](/agentic-studio/overview/security-and-monitoring/internet-access.md)                       |
| Stop paying while my agent is idle                                             | [Auto Pause and Resume](/agentic-studio/overview/sandbox-management/auto-pause-and-resume.md)                           |
| Connect my agent over MCP with one URL                                         | [Sandbox MCP Server](/agentic-studio/overview/integrations-and-support/sandbox-mcp-server.md)                           |
| Save, roll back, or fork sandbox state                                         | [Snapshots](/agentic-studio/overview/sandbox-management/snapshots.md)                                                   |
| Undo an agent's mistake, or duplicate a running sandbox                        | [Undo and Duplicate](/agentic-studio/overview/sandbox-management/undo-and-duplicate.md)                                 |
| See every command an agent ran                                                 | [Audit Trail](/agentic-studio/overview/security-and-monitoring/audit-trail.md)                                          |
| Browse templates through the API                                               | [Sandbox Templates (API)](/agentic-studio/overview/sandbox-management/sandbox-templates.md)                             |
| Monitor CPU, memory, and disk usage                                            | [Metrics](/agentic-studio/overview/security-and-monitoring/metrics.md)                                                  |


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.ai.neevcloud.com/agentic-studio/overview.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.