# Get started

> Takes you from a new org to an agent running in your product, with your coding agent doing the building through Mindset's MCP server.

Building an agent goes in order: build the agent, connect your systems, add functions, scripts and widgets, test it, take it live, then put it in your product.

Mindset runs your agents: the agent loop, the model calls, connections and their credentials, versions, tests and a record of every run. You decide what each agent does and what it can reach. Your coding agent builds it through the `mindset-admin` MCP server, and your app calls it through the SDK or the API.

## Ask your coding agent

```
Connect this project to Mindset's admin MCP server and read how Mindset works from it. Then propose what to build for our invoice exceptions agent: the agent, the connections it needs, and any part that needs a function, script or widget. Build nothing yet.
```

Skill: [mindset-mcp](https://docs.mindset.ai/skills/mindset-mcp/SKILL.md)

## The steps

### 1. Connect your coding agent

Add the `mindset-admin` MCP server and the Mindset skills to your project, and sign in as an org admin. See [Start with your coding agent](https://docs.mindset.ai/docs/work-from-claude-code).

### 2. Keep your agents in your repo

Each agent is a folder of files, reviewed in pull requests and applied to Mindset as versions. See [Keep your agents in your code](https://docs.mindset.ai/docs/keep-your-agents-in-your-code).

### 3. Decide what to build

Most of an agent is prompt text and connections to your systems. Add a function, a script or a widget only where the agent needs one. See [Choose what to build](https://docs.mindset.ai/docs/choose-what-to-build).

### 4. Build it in Sandbox

Build the agent, then [connect your systems](https://docs.mindset.ai/docs/connect-a-system). Every org starts with one environment, **Sandbox**. If the agent already exists in another framework, [move it](https://docs.mindset.ai/docs/move-an-agent-from-another-framework) instead.

### 5. Test it

Run the test parameters of its script and functions, then its behavior tests. See [Test before it goes live](https://docs.mindset.ai/docs/test-before-it-goes-live).

### 6. Move it to Production

Apply the same files to Production. See [Move it to Production](https://docs.mindset.ai/docs/move-it-to-production).

### 7. Put it in your product

[Put the agent on your page](https://docs.mindset.ai/docs/put-an-agent-on-your-page), or [start it from your server](https://docs.mindset.ai/docs/start-it-from-your-server) with no person in a chat.

Creating the org, entering credentials, approving writes and activating a live agent stay with a person. [Work with the MCP server](https://docs.mindset.ai/docs/work-with-the-mcp-server) lists each one.

## Done when

- Your coding agent names the org and environment it's in.
- One agent is kept as files in your repo, tested in Sandbox, and called from your product.