# Test field service changes in a sandbox copy, then promote them to live

> Nobody wants a half-built dispatch rule texting real customers. For anything bigger than a small change, an AI assistant can work in a private copy of your workspace that cannot email, text or call anyone. When it is right, you see exactly what going live would change before you agree.

Source: https://www.fieldproxy.ai/developers/guides/sandbox-to-live

## When a service business uses it

- Reworking the job flow or the technician's app before the busy season, without touching what dispatchers use today.
- Trying new reminder, on-my-way or no-show rules against real jobs without a customer receiving anything.
- Setting up a new branch or line of business alongside the existing one.
- Letting an AI assistant make a large change while the live workspace carries on.

## 1. Make the copy

`create_sandbox` makes a private copy of the workspace to build and test in. In the copy, automations and mailbox sync are off and it can send no email, texts or calls. It is deleted after 7 days unless you extend it (`extend_sandbox`).

- Making the copy disconnects everyone using the live workspace for a few seconds, so the assistant asks first. Pick a quiet moment, not the morning dispatch.
- One sandbox at a time, and one new sandbox a day.
- You get a new MCP address and key for the copy. Add it to your AI client as a second connection.

## 2. Build and test in it

Everything in the other build guides works the same way inside the copy: new tables, apps, automations and agents. Because the copy cannot send anything, you can run a dispatch rule or try a booking agent on your real jobs and customers without one of them hearing about it.

## 3. See what going live would do

From the live workspace's connection, `plan_promotion` works out what copying your chosen apps, automations and agents across would do, without changing anything. It returns the tables and columns live would gain, what each item becomes on live, and anything that blocks it, such as something that points at a user only the sandbox has.

example request:

```text
Plan promoting the subcontractor compliance app and the job sign-off automation
from the sandbox to live. Show me what live would gain before you do anything.
```

## 4. Go live in two safe steps

- `confirm_promotion`, after you agree, adds the tables first, then saves apps as draft versions and automations and agents switched off.
- You then publish each app (`publish_app`) and switch each automation or agent on. Nothing a customer, dispatcher or technician can see changes until you do.

## Questions

### Does Fieldproxy have a sandbox for testing field service changes?

Yes. A workspace can make a private copy of itself to build and test in. The copy cannot send email, texts or calls, its automations are off, and it is deleted after 7 days unless extended.

### How do changes move from the sandbox to the live workspace?

By promotion. You choose the apps, automations and agents to copy, see what live would gain and anything that blocks it, and confirm. Apps arrive as drafts and automations and agents arrive switched off, so you publish and enable them yourself.

### Can an AI assistant disrupt dispatch while it is building?

Building in a sandbox keeps the work away from live entirely. On live, apps are saved as drafts, automations and agents are saved switched off, and data changes are proposed for a person to confirm.

## This is one way, not the only way

This guide shows one common way to do it. A Fieldproxy workspace can be reached from every side, and anything in it can be read, changed or automated:

- REST API: read, insert, update and delete on any table the key allows ([REST API reference](https://www.fieldproxy.ai/developers/docs/rest-api)).
- MCP server: 170+ tools for Claude, ChatGPT, Copilot, Cursor and your own agents, including read-only SQL, proposed writes, automations and app building ([MCP tools](https://www.fieldproxy.ai/developers/docs/mcp-tools)).
- Automations: 15 trigger types and about 50 step types, with webhooks in and signed webhooks and HTTP calls out ([Webhooks and automations](https://www.fieldproxy.ai/developers/docs/webhooks-automations)).
- Public forms, record links and customer portals for people outside the company ([Public forms and links](https://www.fieldproxy.ai/developers/docs/public-forms)).
- Apps as JSON: new screens, fields, tables and workflows ([Apps as JSON](https://www.fieldproxy.ai/developers/docs/apps-as-json)).
- Voice and text agents that work from the same data ([Voice and text agents](https://www.fieldproxy.ai/developers/docs/agents)).
- Not listed here? Fieldproxy's team builds integrations with the other systems you use as part of setting up your workspace. [Tell us what you need](https://www.fieldproxy.ai/book-demo?source=dev-guides).

More: [REST API reference](https://www.fieldproxy.ai/developers/docs/rest-api) · [OpenAPI](https://www.fieldproxy.ai/developers/openapi.json) · [All guides](https://www.fieldproxy.ai/developers/guides)