> For the complete documentation index, see [llms.txt](https://integrations.drivetrain.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://integrations.drivetrain.ai/extending-drivetrain/using-the-drivetrain-mcp-server/connect-common-mcp-clients.md).

# Connect Common MCP Clients

The **Drivetrain MCP server** can be connected to **MCP-compatible AI clients** such as ChatGPT and Claude. Once connected, these clients can retrieve Drivetrain metrics using natural-language prompts.

This guide explains how to connect commonly used MCP clients to your Drivetrain workspace.

### Requirements

Before connecting an MCP client to Drivetrain, ensure the following:

* You have access to your **Drivetrain tenant**
* You know your **tenant subdomain** (for example, in **`acme.drivetrain.ai`**, **`acme`** is the subdomain)
* You are **added as an admin in Drivetrain**
* Your AI platform supports **MCP connectors** (for example, ChatGPT or Claude)

Example MCP endpoint format:

```
https://[InsertTenantSubdomain].drivetrain.ai/drive/api/v1/public/atlas/sse
```

Example:

If your tenant subdomain is **acme**, your endpoint will be:

```
https://acme.drivetrain.ai/drive/api/v1/public/atlas/sse
```

### Connecting to ChatGPT

#### Installation by an Admin

{% hint style="warning" %}
The **ChatGPT administrator** must be added as a **user on your Drivetrain tenant**. This user can be added without a specific role.
{% endhint %}

1. Go to the **ChatGPT web interface**: <https://chatgpt.com>
2. Open the menu in the **bottom-left corner** and navigate to\
   **Settings → Apps → Advanced Settings**.
3. Enable **Developer Mode**.
4. Select **Create app**.
5. Enter a **name and description** for the connector (for example, *Drivetrain*).
6. In the **MCP Server URL** field, paste your Drivetrain MCP endpoint.

Example:

```
https://[InsertTenantSubdomain].drivetrain.ai/drive/api/v1/public/atlas/sse
```

Example with tenant:

```
https://acme.drivetrain.ai/drive/api/v1/public/atlas/sse
```

7. Keep **OAuth** selected as the authentication method.
8. Check the box confirming **“I understand and want to continue.”**
9. Click **Create**.
10. Complete the authentication workflow using your **Drivetrain credentials**.

**Enable the Connector for the Organization**

After creating the connector, publish it for your ChatGPT workspace.

1. Go to **Settings → Apps**.
2. Select the **Drivetrain** app.
3. Click the **three-dot menu**.
4. Select **Publish**.
5. Review the warning message and check **I understand**.
6. Confirm to publish the app for **internal ChatGPT workspace members**.

#### Installation by a User

Once the connector is published, users in the workspace can connect to it.

1. Ensure you are added as a **user in Drivetrain**:\
   `https://[InsertTenantSubdomain].drivetrain.ai`
2. In ChatGPT, open the **Apps** section in the left sidebar.
3. Locate the **Drivetrain** app published by your administrator.
4. Click **Connect**.
5. Authenticate using your **Drivetrain credentials**.

To use the connector:

1. Start a **new chat**.
2. Click the **+ icon** in the prompt bar.
3. Select the **Drivetrain connector**.
4. Enter your query.

The client can now retrieve Drivetrain metrics through MCP tools.

### Connecting to Claude (OAuth)

#### Installation

The user installing the connector must have **Owner privileges** in the Claude workspace.

1. Go to the **Claude web interface**: <https://claude.ai>
2. Click the **profile icon** in the bottom-left corner.
3. Open **Admin settings**.
4. Select the **Connectors** tab.
5. Click **Add custom connector**.
6. Enter a **name** for the connector.
7. In the **Remote MCP server URL** field, paste your Drivetrain MCP endpoint.

Example:

```
https://[InsertTenantSubdomain].drivetrain.ai/drive/api/v1/public/atlas/sse
```

8. Leave **OAuth Client ID** and **OAuth Client Secret** empty.
9. Click **Add**.
10. Click **Connect** and complete authentication using your Drivetrain credentials.

The connection is successful once the connector status shows **Configure**.

**Note:** Claude may move connectors to the **Customize** section in future updates. If this occurs, navigate to:

```
Customize → Connectors → + → Add Custom Connector
```

Then follow the same configuration steps.

### Connecting to Claude (API Key)

{% hint style="info" %}
You must be an **admin in Drivetrain** to complete this connection.
{% endhint %}

To connect the **Claude Desktop App with Drivetrain**, follow these steps:

**Obtain your credentials**

You will need your **API key** from Drivetrain.

1. In Drivetrain, go to **Settings → API Key.**
2. **Copy** the API key shown.

**Edit the client configuration**

Next, update your **MCP client** to include **Drivetrain**.

1. In Claude Desktop, go to **Settings → Developer → Edit Config**.
2. **Right-click** and open the config file in **any text editor** (for example, Sublime Text or Notepad).
3. **Copy** **and** **paste** the **Drivetrain MCP server snippet** below into the **config file** based on your machine (Windows or Mac).

{% tabs %}
{% tab title="For PC / Windows Machines" %}

```json
{
  "mcpServers": {
    "Drivetrain": {
      "command": "cmd.exe",
      "args": [
        "/C",
        "C:\\PROGRA~1\\nodejs\\npx.cmd",
        "-y",
        "mcp-remote",
        "https://[InsertTenantSubdomain].drivetrain.ai/drive/api/v1/public/atlas/sse",
        "--header",
        "apikey:API<>KEY"
      ]
    }
  }
}
```

{% endtab %}

{% tab title="For Mac" %}

```json
{
"mcpServers": {
    "drivetrain": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        
        "https://[InsertTenantSubdomain].drivetrain.ai/drive/api/v1/public/atlas/sse",
        "--header",
        "apikey: <YOUR_API_KEY>"
      ]
    }
  }
}
```

{% endtab %}
{% endtabs %}

**Replace the placeholders**

**Before saving** the file, make the following substitutions:

1. **\[InsertTenantSubdomain]** → replace this with your **company’s Drivetrain tenant subdomain** (e.g., acme, microsoft, etc.).
2. **\<YOUR\_API\_KEY>** → replace this with the **API key you copied** in step 1.

**Example**

If your Drivetrain tenant subdomain is **"acme"** and your API key is **"abcd1234"**, the config would look like this:

```json
{
"mcpServers": {
    "drivetrain": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
"https://acme.drivetrain.ai/drive/api/v1/public/atlas/sse",
        "--header",
        "apikey: abcd1234"
      ]
    }
  }
}
```

#### Restart the client

1. **Save** the config file and **restart Claude Desktop.**
2. Once **restarted**, Drivetrain will appear as a **connected MCP server.**

#### Troubleshooting

* **ENOENT** **error** during setup: This error may appear if a required dependency is missing. Install the provided package and restart the client to resolve the issue. (<https://nodejs.org/en/download>)
* **Invalid credentials:** Verify that your API key is copied correctly from Drivetrain.

### Testing the connection

After connecting the MCP client, open a new chat and try a simple query.

Example prompts:

```
Show my Revenue for the last 12 months
```

```
What is my Revenue per FTE?
```

If the connection is successful:

* The AI client will detect **Drivetrain MCP tools**
* The client will retrieve **metric values from your Drivetrain workspace**
* Results will appear directly in the chat response

If the client cannot access Drivetrain, verify that:

* The **MCP endpoint URL** is correct
* You are **authenticated with Drivetrain**
* The connector is **enabled in your AI client workspace**


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://integrations.drivetrain.ai/extending-drivetrain/using-the-drivetrain-mcp-server/connect-common-mcp-clients.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.