---
updatedAt: 2026-07-21T13:34:47.000Z
agentTools:
  projectIndex: https://developer.remote.com/llms.txt
---

# Sandbox Quickstart

Get a seeded sandbox account and a working API token, then confirm the token can reach the API. That's it!

## Before you start

* **Sandbox and production are fully isolated.** They use separate accounts and separate tokens, so nothing you do in the sandbox touches production data.
* **Your token is a Customer API token.** Sandbox tokens are prefixed `ra_test_`; production tokens are prefixed `ra_live_`. If your token starts with `ra_live_`, stop — that is a production token and will not authenticate against the sandbox.

## Step 1: Get a sandbox and token

Request a seeded sandbox through the sandbox request form: [Open request form](https://remote.com/remote-mcp/#get-started)

1. Submit your email address on the sandbox request form.
2. Open the confirmation email and click the verification link.
3. Remote hands you a pre-built demo company and emails you its login credentials and a `ra_test_` API token.

The demo company is seeded with sample data and it stays available for 14 days before access is revoked.

## Step 2: Verify the token

Set your token, then call the identity endpoint. A `200` confirms the token works and shows the entities it controls.

```bash
export REMOTE_API_TOKEN=ra_test_...

curl https://gateway.remote-sandbox.com/v1/identity/current \
  -H "Authorization: Bearer $REMOTE_API_TOKEN"
```

A `200` response means you are ready. For a Customer API token, the response body identifies the company and the user this token can act on behalf of.

## Next steps

* [Your first API call](https://developer.remote.com/docs/your-first-api-call)
* [Introduction to the Remote CLI](https://developer.remote.com/docs/introduction-to-the-remote-cli)
* [Remote CLI quickstart](https://developer.remote.com/docs/remote-cli-quick-start)
* [Configure your AI Agent to build for you](https://developer.remote.com/docs/build-integrations-with-ai)

<br />