- AI Productivity Insights
- Pages
- Step 4: Connect Calude Desktop to the MCP Server
Step-by-Step Guide: Creating an MCP Server with Azure AI Agent Service for TripAdvisor Marketing Campaign
Step 4: Connect Calude Desktop to the MCP Server
Open the Claude Desktop App, click the hamburger menu, select File, then go to Settings.

Select the Developer menu and click "Edit Config".

Select the file name:
claude_desktop_config.json
and double click on the file to open it.

Copy the configuration file, your project connection string, your agent id and past it into the file:
claude_desktop_config.json
and save it.
{
"mcpServers": {
"azure-agent": {
"command": "C:\\mcp-foundry\\.venv\\Scripts\\python.exe",
"args": [
"-m",
"azure_agent_mcp_server"
],
"cwd": "C:\\mcp-foundry\\src\\python",
"env": {
"PYTHONPATH": "C:\\mcp-foundry\\src\\python",
"PROJECT_CONNECTION_STRING": "YOUR_PROJECT_CONNECTION_STRING",
"DEFAULT_AGENT_ID": "YOUR_AGENT_ID"
}
}
}
}

Open the Claude Desktop App, click the hamburger menu, select "File," then go to "Settings." Select "Developer" and choose
azure-agent
. If successful, you will see a "running" status.

Exit the Developer menu. You will then see a hammer icon with "
3 MCP tools available
" in the chat box.

Click on the hammer icon to see the available MCP tools of Azure AI Agent.
