After deploying your Semantic Model MCP Endpoint, you will configure the MCP Server by connecting it to your Prepare instance database and creating one or more MCP services. Each service exposes a specific semantic model on its own port with unique credentials.
Prerequisites
Install TimeXtender MCP Server
- Complete the Early Access Sign-up
- Download the TimeXtender MCP Installer
- Run the installer executable
- Follow the installation wizard prompts
- Wait for the TimeXtender MCP Configurator to launch automatically upon installation completion
Configure Prepare Instance Connection
- Navigate to the Prepare Instance Config tab in the TimeXtender MCP Configurator
- Enter the Server connection string for your Prepare instance database
- Example: myserver.database.windows.net
- Enter the Database name
- Example: mydb
- Enter the User ID for database authentication
- Use a read-only database user account for security
- While the MCP server does not support ALTER, UPDATE, or DELETE statements, AI behavior can be unpredictable. Read-only credentials ensure data warehouse reliability.
- Enter the Password for the database user
- Click Test Connection to verify connectivity
- Click Save once the connection test succeeds
Add MCP Service
TimeXtender supports creating multiple MCP Services, One for each Semantic Model.
- Click the Service Management tab in the TimeXtender MCP Configurator
- Click Add Service to open the service configuration dialog
- Enter the Port number for the MCP service
- Default range: 6000-6100
- Test the port before proceeding
- Paste the full file path of the semantic model JSON file in the Semantic Layer File field
- Use the same path specified when creating the deliver instance
- Example: C:\TimeXtender\sales.json
- Enter a descriptive Service Name or use the auto-generated name
- Select the Startup Type (Automatic recommended)
- Select the Service Account (Network Service recommended)
- Copy the API Key from the configuration dialog
- Store the API key securely for AI agent authentication
- Copy the Remote MCP Server URL
- Use this URL to configure client-server connections in tools such as LM Studio
- Format: http://[server-ip]:[port]/
- Click Save to create the MCP service
You may repeat this process for each semantic model using different Semantic Layer files and Port numbers
Troubleshooting
Invalid JSON file path error
Verify the JSON file path uses only letters, numbers, hyphens (-), and underscores (_). Remove spaces and special characters. Ensure the filename uses the .json extension.
MCP service fails to start
Check that the specified port is not in use by another application. Test the port in the Add MCP Service dialog before saving. Verify the semantic layer file exists at the specified path and contains valid JSON.
Connection timeout when testing database connection
Increase the Connection timeout value in the Prepare Instance Config tab. Verify firewall rules allow connections to the database server. Confirm the database server address and credentials are correct.
