Overview
Adpulse is backed by a GraphQL API that supports all the data and functionality available in the app.
We have created an MCP server that can access this API directly, with Connectors for Claude, ChatGPT, and Gemini.
How to generate the MCP credentials from within Adpulse
Navigate to: Settings > Integrations > MCP Servers
Click 'Add Server'
Choose your AI from the list (Claude, ChatGPT, Gemini, Custom) - this is important as some have different requirements.
This will create a row with the below fields;
Remote MCP server URL
OAuth Client ID
OAuth Client Secret
You will need to enter these three fields into your AI Client to create the connection.
How to connect Claude
Currently, you'll need to connect via a Custom Connector in Claude.
1. Navigate to the Custom Connector option
Customize > Connectors > Add custom connector
2. Add Custom Connector
You will need to source these fields from Adpulse, as outlined above.
Name - Free text, and we'd recommend calling it 'Adpulse'
Remote MCP server URL - from Adpulse
3. Click Add
How to connect ChatGPT
Currently, you'll need to connect by creating a new App within ChatGPT - this will require you to enable 'Developer Mode'.
1. Navigate to the Apps page within ChatGPT
2. Enable Developer Mode
To create a custom App, ChatGPT requires you to enable Developer Mode:
Click the Settings cog at the top right of the page
Click on Advanced Settings
Toggle on Developer Mode
3. Create an App via the new button now available
ChatGPT offers many more options, but you still only need the three fields provided by Adpulse.
You will need to source these fields from Adpulse, as outlined above.
Name - Free text, and we'd recommend calling it 'Adpulse'
Description - Free text and optional
Connection - the Remote MCP server URL from Adpulse
Open the Advanced OAuth settings
Advanced OAuth Settings
This will open a side panel with many options, but we only need the top one where you will enter:
Once you enter them and accept the terms, you can create the App.
How to connect Gemini
You can connect Gemini in multiple ways; the below outlines how to connect Gemini Enterprise via Google Cloud and Gemini Enterprise via the Gemini web app. Both use the same details, but they are set up in different places.
Gemini requires some extra fields to be populated; see below the full list you'll need:
Required fields
Field | Value |
Authorization URL | |
Token URL | |
Client ID | |
Client Secret | |
Scopes | offline_access |
Gemini Enterprise via Google Cloud
1. Navigate to Gemini Enterprise with your Google Cloud Project
2. Then Data Sources > [Create new Data Source]
3. In the options, you'll have [Add MCP Server]
4. You will be presented with a page where you enter the details from above
5.. Ensure both of these are checked:
6. Click [Verify auth] once all the details are completed to ensure the connection will connect correctly. This will open an authorization window, so follow the usual login prompts.
7. Once verified, click [Continue]
8. The final step is to:
a. Give the connector a name
b. Assign your Sensitive data protection policy
c. and Configure access control
Both B and C above will be unique to your organization.
Gemini Business Web App
1. Navigate to Gemini Enterprise
2. Navigate to the admin settings; click on the settings cog > your team > manage team
3. Navigate to Connected apps and click [Add MCP Server]
4.. In the next two modals, you'll enter the values from above and click [Add]
5. You'll now see the connector in the list:
6. Next, you'll need to connect to the MCP Server, so navigate back to the standard chat interface
7. Navigate to the Connector settings; click on the settings cog > Connectors
8. This will open up a list of the available connectors; click [Connect] next to the one you just created. You'll log in using your credentials, and once you're done, you'll have it connected.
In Gemini, you need to authorize a specific chat to use the Adpulse MCP, you can do that here:
