• Prerequisites
• Create a Connector
• Create an Agent and Exercise the Connector

Prerequisites

Generate a Personal Access Token (PAT) on the Settings page of CData Connect Cloud. Copy this down, as it acts as your password during authentication.

Create a Connector

1. Copy and save the following code as an OpenAPI file:

    swagger: '2.0'
    info:
    title: QueryMCP
    description: Connect Cloud Query MCP Copilot Integration
    version: 1.0.5
    host: cloud-16284.eastus-dev.mcp.clouddataos.com
    basePath: /
    schemes:
    - https
    paths:
    /mcp:
        post:
        summary: Query MCP Server
        x-ms-agentic-protocol: mcp-streamable-1.0
        operationId: InvokeMCP
        parameters:
            - name: Accept
            in: header
            required: true
            type: string
            default: application/json, text/event-stream
            description: Content types that the client can handle
        responses:
            '200':
            description: Success
    securityDefinitions:
    basic-auth:
        type: basic
    security:
    - basic-auth: []
   

2. Log in to Microsoft Copilot Studio.

3. In the navigation menu, select Tools, and then click the New Tool button.

4. In the New tool dialog, click Custom connector. You are redirected to Microsoft Power Apps.

5. Select New custom connector > Import an OpenAPI file.

   

6. Select the file you saved in step 1 and click Continue.

7. In the 1. General step, enter general information about the connector, including a description and a new icon (optional). Host and Base URL are required.

   • Host–enter mcp.cloud.cdata.com.

   • Base URL–enter a forward slash (/).

   

8. In the 2. Security step, the Authentication type should be Basic authentication. Ensure that the Parameter labels are set to username and password.

   

9. Leave 3. Definition and 4. Code as is.

10. Click Create connector. You should now have a testable connector.

11. In the 5. Test step, click +New connection. Click Create.

12. Enter your username and password. The username is your CData Connect Cloud username, and the password is the PAT you created in the prerequisites.

13. Click Save. You may have to find the connector again in the Custom connectors screen and click Edit.

14. Return to 5.Test. Select the new connection you just created.

15. Enter the following JSON snippet in InvokeMCP (turn on Raw Body):

    {
      "jsonrpc": "2.0",
      "id": 1,
      "method": "tools/call",
      "params": {
         "name": "getCatalogs",
         "arguments": {}
      }
    }
    

    

16. Click Test operation. You now have a live connector and can return to Copilot Studio.

Create an Agent and Exercise the Connector

1. In the Copilot Studio navigation menu, click Agents, and then click +New agent.

2. Toggle the Configure button.

3. Enter information based on how you intend to use this agent while it is consuming our MCP server.

4. Click Create on the upper right. Copilot shows you the agent details.

   

5. Under Tools, click Add tool. Limit the selection to Model Context Protocol and click the Query MCP Server connector you created.

   

6. Click Add to agent to add your tool.

7. You can now exercise this connector. Select the Topics tab and click +Add a topic.

   

8. Enter a question in the chat window. Since this is the first time you are exercising this connector, Copilot displays a message telling you to open the connection manager. Click the link.

   

9. Select the connection you created and click Connect. In the Create or pick a connection dialog, click Submit.

10. Return to the Agent window and click Retry to retry your request.