> For the complete documentation index, see [llms.txt](https://docs.datalogz.io/guides/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.datalogz.io/guides/datalogz-mcp-servers/connecting-to-an-mcp-server.md).

# Connecting to an MCP Server

The Control Tower MCP server speaks the standard Model Context Protocol, so any compliant client can connect to it (Claude Desktop, Claude Code, Cursor, VS Code, or your own agent built on an MCP SDK).

#### Before you start

* **At least one connector onboarded and scanned.** The server reports on data Control Tower has already collected; it does not scan anything itself. See the Connector Onboarding Guide.
* **A Control Tower credential.** The agent inherits that credential's connector scope and it sees exactly what the credential is entitled to see, and nothing else.
* **Admin access to Claude.** You need to be able to add new connectors to Claude.

Connect your AI client to Control Tower with one URL:

```
https://app.datalogz.io/mcp
```

Use the HTTPS URL supplied by Datalogz. You do not need a client ID, client secret, or API token. When prompted, sign in with a verified Control Tower administrator account and approve access.

### Claude Desktop

1. Open **Customize > Connectors**.
2. Select **Add custom connector** and enter the Control Tower MCP URL.
3. Leave the advanced OAuth client ID and secret fields empty, then select **Add**.
4. Select **Connect** and complete the browser sign-in.

For Team and Enterprise plans, an organization owner must first add the connector from **Organization settings > Connectors > Add > Custom > Web**. Members can then connect it from **Customize > Connectors**.

Enable Control Tower for a conversation from **+ > Connectors**.

### Claude Code

Add Control Tower for all your projects:

```bash
claude mcp add --transport http --scope user \
  control-tower "https://app.datalogz.io/mcp"
```

In Claude Code, run `/mcp`, select `control-tower`, and complete the browser sign-in.

### ChatGPT desktop app

In newer app versions, MCP management is not shown in the Settings sidebar:

1. From the main app open the command menu with `Command+K` on macOS or `Ctrl+K` on Windows and Linux.
2. Search for `MCP` and select **MCP Servers**.
3. Select **Add Server**.
4. Name it `Control Tower`, choose **Streamable HTTP**, enter the Control Tower MCP URL, `https://app.datalogz.io/mcp` and select **Save**.
5. Select **Authenticate**, complete the browser sign-in, and restart the app if prompted.

If the MCP command is unavailable, configure the shared MCP connection from a terminal and then restart the app (see "Codex" below)

### Codex

Add the server and start OAuth sign-in:

```bash
codex mcp add control-tower --url "https://app.datalogz.io/mcp"
codex mcp login control-tower
```

Skip these commands if you already added the server for the ChatGPT desktop app.

Run `codex mcp list`, or use `/mcp` in the Codex terminal UI, to confirm the connection. Codex and the ChatGPT desktop app share MCP configuration on the same computer.

### Microsoft Copilot Studio

1. Open your agent's **Tools** page and select **Add a tool > New tool > Model Context Protocol**.
2. Enter a name, a short description, and the Control Tower MCP URL.
3. For authentication, select **OAuth 2.0 > Dynamic discovery**.
4. Select **Create**, then **Next**.
5. Create a new connection, complete the browser sign-in, and select **Add to agent**.

### Manually create a client

In Datalogz Control Tower:

1. Go to **Settings** > **OAuth Clients**.
2. Click **New OAuth Client**.
3. Fill out the following fields:

<figure><img src="https://1435452537-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FI51V8nP3ZjMKt2bCO9HR%2Fuploads%2FaYycD5wjxIW2iRNbreOv%2FScreenshot%202026-08-20%20at%2016.03.58.png?alt=media&amp;token=fa40487f-eb44-4f91-b698-465bb39a6072" alt="" width="375"><figcaption></figcaption></figure>

<table><thead><tr><th width="262.61328125">Value</th><th>Description</th></tr></thead><tbody><tr><td><strong>Name</strong></td><td>The name you choose for the Client.</td></tr><tr><td><strong>Type</strong></td><td>You can create two types of OAuth clients. For the MCP oAuth client, you select the <strong>Interactive PKCE</strong> type.</td></tr><tr><td><strong>Redirect URIs</strong></td><td>Select or fill out the callback URIs. If you want to add the MCP server to Claude or Copilot, you can select the predefined URIs.</td></tr><tr><td><strong>Advanced - Connector access</strong></td><td>Limit the scope of the MCP server by linking it to a specific connector. Leave this part blank to allow the tool to search every connector your credential can reach; pass one and the tool answers for that connector alone.</td></tr></tbody></table>

4. Click **Create**.
5. Copy the **Client ID**.

In Claude:

1. Go to **Organization Settings**.
2. Click **Connectors** > **+ Add**
3. Select **Custom** and then **Web**.
4. Fill out the following fields:

<figure><img src="https://1435452537-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FI51V8nP3ZjMKt2bCO9HR%2Fuploads%2FGOahfj8gXjGWNKbFF0Y7%2FScreenshot%202026-08-20%20at%2016.20.09.png?alt=media&amp;token=0c79761d-91da-482a-9418-8283d580d4eb" alt="" width="375"><figcaption></figcaption></figure>

<table><thead><tr><th width="262.61328125">Value</th><th>Description</th></tr></thead><tbody><tr><td><strong>Name</strong></td><td>The name you choose for the MCP server.</td></tr><tr><td><strong>Remote MCP server URL</strong></td><td>Type https://app.datalogz.io/mcp</td></tr></tbody></table>

{% embed url="<https://app.arcade.software/share/XeB1Bds0IWBRd1hBn2ES>" %}

### Troubleshooting

* Confirm that the URL uses **HTTPS** and ends in `/mcp`.
* **If authentication expires or is revoked**, reconnect the server and sign in again.
* **If approval is denied**, ask a Control Tower account administrator to complete the connection.
* **If a tool returns nothing at all.** Your credential may have no access to the connector in question.&#x20;
* **If a platform-specific tool refuses your asset.** Usage, viewer, workspace-access and capacity tools are Power BI only. Similarity is Power BI and Tableau. Downstream impact is Power BI, Tableau and Sigma. See the Tool reference.
* **If the numbers look stale.** The server reads what the last connector scan wrote. Check scan status in the app rather than retrying the query.

For client-specific details, see the official setup guides for [Claude connectors](https://claude.com/docs/connectors/custom/remote-mcp), [Claude Code](https://code.claude.com/docs/en/mcp), [ChatGPT and Codex](https://learn.chatgpt.com/docs/extend/mcp), and [Microsoft Copilot Studio](https://learn.microsoft.com/en-us/microsoft-copilot-studio/mcp-add-existing-server-to-agent).


---

# 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://docs.datalogz.io/guides/datalogz-mcp-servers/connecting-to-an-mcp-server.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.