🔹 Feature: Azure App Service as a Model Context Protocol (MCP) Server
🔹 What It Does: Host your web app as an MCP server on Azure App Service and expose its REST APIs as AI-accessible tools that agents (GitHub Copilot Chat, Cursor, Windsurf, etc.) can discover and invoke.
What Is It Giving You:
✅ Turn Existing APIs Into Agent Tools: No rewrite. Existing App Service REST endpoints become MCP tools with structured inputs and outputs.
✅ Discovery + Invocation Flow: The agent queries your MCP server for available tools, picks the right one for the user’s intent, your App Service processes the request, and the agent returns the result to the user.
✅ Built-In Authentication: Native Microsoft Entra ID integration and managed identity support for production-grade MCP servers. Full control over authN/authZ and data access.
✅ Multi-Language Tutorials: End-to-end tutorials for .NET, Java, Node.js, and Python — integrate an App Service app as an MCP server for GitHub Copilot Chat.
✅ Reuse Your Existing Deployment: The same App Service that serves your line-of-business app also serves as the MCP endpoint. One app, one infra, one identity model — used by both humans and agents.
✅ Use Cases That Actually Ship: Developer productivity tools (agents can call your APIs from the IDE), custom copilot experiences for line-of-business apps, multi-step workflow automation, and internal AI-accessible services.
✅ Compatible With Any MCP Client: Not tied to a specific agent. Any MCP-compatible client — GitHub Copilot Chat, Cursor, Windsurf, Claude Code, custom agents — works out of the box.
✅ Secured MCP Calls From VS Code: Dedicated guide for securing MCP calls from Visual Studio Code with Microsoft Entra authentication.
🌐 https://learn.microsoft.com/en-us/azure/app-service/scenario-ai-model-context-protocol-server?tabs=dotnet
