LLM-ready API
Model Context Protocol (MCP)
Third Party Integrations
Microsoft Copilot Studio

Microsoft Copilot Studio MCP Integration

These instructions walk you through integrating the S&P Global LLM-Ready API as a Microsoft Copilot Studio (opens in a new tab) agent, which enables you to use it with Microsoft 365 and Microsoft Teams.

Setup

Step 1: Navigate to Microsoft Copilot Studio

Navigate to Microsoft Copilot Studio (opens in a new tab). On the left sidebar, select Agents.

Copilot page

In the top right corner, select Create blank agent.

Create blank agent

Step 2: Configure your agent

Fill in the details of your agent: the name, description, model, instructions, etc. Scroll to the Tools section and select Add tool.

Add tool

Step 3: Select Model Context Protocol

In the Add tool popup, select Model Context Protocol under the Create new sub-menu.

MCP option

Step 4: Add the LLM-Ready API Model Context Protocol server

Fill in the Server name and Server description, and set the following fields:

  • Server URL: https://kfinance.kensho.com/integrations/mcp
  • Authentication: OAuth 2.0
  • Type: Dynamic discovery

Select Create to add the server. This may take a moment.

MCP fields

Step 5: Create a new connection

Your tool should now be created. Under Connection, select Not connected > Create new connection. This will take you through an Okta browser login flow.

Create new connection

Step 6: Log in with Kensho Okta credentials

Log in with your Kensho Okta credentials, which were sent via email when you signed up for a trial or account. Completing the Okta sign-in will redirect you back to Copilot.

Okta login flow

Step 7: Add and configure the connection

Once you are redirected back to Copilot, select Add and configure.

Add and configure

You should now see a list of available tools.

Tool list

Step 8: Test your agent

In the top left corner, select Test to test your agent. For example, try asking "Balance sheet for XOM 2025Q2". You will be prompted to connect before continuing. Select Open connection manager to open the Manage your connections page. Make sure to keep the existing chat window open.

Test your agent

Step 9: Manage your connections

Find the LLM-Ready API connection you created and select Connect.

Connect to LRA

This will take you through the Microsoft and Kensho Okta authentication flow. Once authenticated, you should see a green checkmark next to your connection. Select Submit.

Submit connection

You should now see that you are connected. You can close this tab.

Connected

Step 10: Retry your connection

Return to your test chat window and select Retry.

Retry connection

You should receive a successful response from the agent with data retrieved via the LLM-Ready API MCP tools, similar to the image below.

Example tool call

Step 11: Connect to your channels

To connect your agent to Microsoft channels such as Microsoft 365 and Microsoft Teams, select Channels > Microsoft 365 and Microsoft Teams at the top of the page.

Channels tab

A panel will open on the right side. Select Add channel in the bottom left corner.

Add channel

In the popup, choose whether to push the newest version, then select Publish.

Publish

Your agent is now available on Microsoft channels. Select See agent in Microsoft 365 or See agent in Teams to use your agent on the corresponding platform.

See agent

Step 12: Use your agent

With your agent and connection set up, the LLM-Ready API is now available in Microsoft 365 and Microsoft Teams. Your agent will appear under the Agents tab in Teams.

Agent in Teams