Overview
This guide explains how to use Guru's Model Context Protocol (MCP) Server to connect your TeamAI Workspace. The integration allows you to access company knowledge directly within supported AI tools using verified information.
Learning Objectives:
Understand what Guru MCP is and its benefits
Set up connections between Guru and MCP-compatible tools
Access and manage knowledge base content within AI tools
Learn specific use cases for different teams
Prerequisites
You'll Need:
Access to your workspace with appropriate permissions
An MCP-compatible AI tool (ChatGPT Enterprise, Claude Desktop, Cursor IDE, etc.)
Authentication credentials for Guru
Understanding Guru MCP
Review how Guru MCP functions with TeamAI
MCP allows TeamAI to search and reference verified company knowledge from Guru
The integration maintains security through Guru's permission system
Information remains up-to-date as it's retrieved directly from your knowledge base
Result: TeamAI can access your company's knowledge directly without requiring manual copying of information.
Note: MCP is an open-source protocol that standardizes how AI models connect to knowledge sources.
Setting Up Guru MCP in TeamAI
Creating the MCP Tool
Navigate to the TeamAI Tools icon located in the left menu
Find and click on the Tools section in the main navigation panel
Click on "Create a new tool"
Look for the create button or plus icon
A new window will appear asking for tool configuration
Configure the basic tool settings
Name: Enter "Guru MCP"
Group: Select the folder where the tool will be stored
Visibility: Choose Workspace, Personal, or Organization based on your needs
Tool Type: Select "MCP Server"
Click "Create"
Result: The tool will be created and you'll be taken to the configuration screen.
Configuring the MCP Connection
Set up tool details
Select an avatar icon or upload a custom image
Update the folder (group) if needed
Navigate to Server Setup
Click on "Server setup" in the left menu below details
Enter the Guru MCP endpoint
In the MCP server URL field, enter: https://mcp.api.getguru.com/mcp
In the authentication section, select "OAuth 2.1 Authentication"
Complete the authorization process
Click on "Connect with auth"
A new window will open for authorization
Click "Grant permission" to authorize the connection
Verify the connection
Check for a confirmation message
Ensure the tool status shows as connected
Result: The Guru MCP tool will be configured and ready to use in your TeamAI workspace.
Important: The connection uses your personal Guru credentials, so you'll only have access to Guru content that you're authorized to view.
Using Guru MCP in TeamAI
Accessing Guru Knowledge
Start a new conversation in TeamAI
Open the chat interface
Activate the Tool recently created on the right panel by clicking on the toggle
Ask questions about company information
Formulate queries that would be answered by your Guru knowledge base
Example: "What is our company's return policy?"
Review the responses
Verify that answers include citations to Guru content
Check source information for accuracy
Result: TeamAI will provide information from your Guru knowledge base with proper attribution.
Best Practices
Use specific queries: Frame questions clearly to get the most relevant knowledge base results.
Verify information sources: Always check the citations provided with responses.
Respect permissions: Remember that the tool only provides access to Guru content you're authorized to view.
Report inaccuracies: If you discover outdated information, update it in Guru to ensure future responses are accurate.
Common Questions
Q: Can I connect to multiple Guru knowledge bases? A: You would need to create separate MCP tools for each knowledge base you want to connect.
Q: What happens if information in Guru changes? A: The integration accesses your live Guru knowledge base, so information is as current as your knowledge base content.
Q: Can other users in my workspace use my Guru MCP connection? A: If you set the visibility to Workspace or Organization, others can use the tool, but they'll only see content they have permission to access in Guru.
Q: Do I need to reconnect the tool regularly? A: You may need to reauthorize the connection if your Guru credentials change or if the authorization expires.
Q: What if the connection fails? A: Verify your Guru credentials are current, check your network connectivity, and try reconnecting through the tool settings.
Q: Can I customize what information is accessible? A: The integration respects Guru's permission system, so customization should be done within Guru.