draft

dlepow · dlepow · commit 36c85eadb0f0 · 2025-05-12T13:46:34.000-07:00
diff --git a/articles/api-center/includes/about-mcp-servers.md b/articles/api-center/includes/about-mcp-servers.md
@@ -0,0 +1,43 @@
+---
+title: Include file
+description: Include file
+services: api-center
+author: dlepow
+
+ms.service: azure-api-center
+ms.topic: include
+ms.date: 05/12/2025
+ms.author: danlep
+ms.custom: Include file
+---
+
+## About MCP servers
+
+AI agents are becoming widely adopted because of enhanced large language model (LLM) capabilities. However, even the most advanced models face limitations because of their isolation from external data. Each new data source potentially requires custom implementations to extract, prepare, and make data accessible for the models.
+
+The [model context protocol](https://www.anthropic.com/news/model-context-protocol) (MCP) helps solve this problem. MCP is an open standard for connecting AI models and agents with external data sources such as local data sources (databases or computer files) or remote services (systems available over the internet, such as remote databases or APIs).
+
+### MCP architecture
+
+The following diagram illustrates the MCP architecture:
+ 
+:::image type="content" source="media/about-mcp-servers/mcp-architecture.png" alt-text="Diagram of model context protocol (MCP) architecture.":::
+
+The architecture consists of the following components:
+
+| Component      | Description                                                                                     |
+|----------------|-------------------------------------------------------------------------------------------------|
+| **MCP hosts**  | LLM applications such as chat apps or AI assistants in your IDEs (like GitHub Copilot in Visual Studio Code) that need to access external capabilities |
+| **MCP clients**| Protocol clients, inside the host application, that maintain 1:1 connections with servers        |
+| **MCP servers**| Lightweight programs that each expose specific capabilities and provide context, tools, and prompts to clients |
+| **MCP protocol**| Transport layer in the middle        |
+
+MCP follows a client-server architecture where a host application can connect to multiple servers. Whenever your MCP host or client needs a tool, it connects to the MCP server. The MCP server then connects to, for example, a database or an API. MCP hosts and servers connect with each other through the MCP protocol.
+
+### Remote versus local MCP servers
+
+MCP utilizes a client-host-server architecture built on [JSON-RPC 2.0 for messaging](https://modelcontextprotocol.io/docs/concepts/architecture). Communication between clients and servers occurs over defined transport layers, and supports primarily two modes of operation:
+
+* **Remote MCP servers** - MCP clients connect to MCP servers over the internet, establishing a connection using HTTP and server-sent events (SSE), and authorizing the MCP client access to resources on the user's account using OAuth.
+
+* **Local MCP servers** MCP clients connect to MCP servers on the same machine, using standard input/output as a local transport method.
diff --git a/articles/api-center/includes/api-center-portal-compare-apim-dev-portal.md b/articles/api-center/includes/api-center-portal-compare-apim-dev-portal.md
@@ -6,31 +6,38 @@ author: dlepow
 
 ms.service: azure-api-center
 ms.topic: include
-ms.date: 03/05/2025
+ms.date: 05/12/2025
 ms.author: danlep
 ms.custom: Include file
 ---
 
-## API Management and API Center portals
+## About MCP servers
 
-The [Azure API Management](../../api-management/api-management-key-concepts.md) and [Azure API Center](../overview.md) services both provide portals for developers to discover and consume APIs:
+AI agents are becoming widely adopted because of enhanced large language model (LLM) capabilities. However, even the most advanced models face limitations because of their isolation from external data. Each new data source potentially requires custom implementations to extract, prepare, and make data accessible for the models.
 
-* The *API Management developer portal* allows users to find managed APIs, learn how to use them, request access, and test them.
-* The *API Center portal* (preview) is a multi-gateway portal where  users can discover and filter the organization's complete API inventory.
+The [model context protocol](https://www.anthropic.com/news/model-context-protocol) (MCP) helps solve this problem. MCP is an open standard for connecting AI models and agents with external data sources such as local data sources (databases or computer files) or remote services (systems available over the internet, such as remote databases or APIs).
 
-While the two portals share some features, they also have distinct differences. The following table compares current capabilities to help determine which portal to use. Some organizations may prefer one portal, while others may need both.
+### MCP architecture
 
-| Feature | API Management developer portal | API Center portal (preview) |
-| --- | --- | --- |
-| Search and filter API inventory | API Management instance only | All APIs<sup>1</sup> |
-| View API details and definitions | ✔️ | ✔️ |
-| View API documentation | ✔️ | ✔️ |
-| Customize with branding | ✔️ | Name only |
-| Integrate with Microsoft Entra ID | ✔️ | ✔️ |
-| Add custom widgets | ✔️ | ❌ |
-| Customize with WordPress | ✔️ | ❌ |
-| Test APIs in test console | ✔️ | ✔️ |
-| Subscribe to APIs | ✔️ | ❌ |
-| View API usage analytics | ✔️ | ❌ |
+The following diagram illustrates the MCP architecture:
+ 
+:::image type="content" source="media/register-discover-mcp-server/mcp-architecture.png" alt-text="Diagram of model context protocol (MCP) architecture.":::
 
-<sup>1</sup> The API Center portal can contain all APIs in your organization, including those managed in Azure API Management and other platforms, as well as unmanaged APIs and APIs under development.
+The architecture consists of the following components:
+
+| Component      | Description                                                                                     |
+|----------------|-------------------------------------------------------------------------------------------------|
+| **MCP hosts**  | LLM applications such as chat apps or AI assistants in your IDEs (like GitHub Copilot in Visual Studio Code) that need to access external capabilities |
+| **MCP clients**| Protocol clients, inside the host application, that maintain 1:1 connections with servers        |
+| **MCP servers**| Lightweight programs that each expose specific capabilities and provide context, tools, and prompts to clients |
+| **MCP protocol**| Transport layer in the middle        |
+
+MCP follows a client-server architecture where a host application can connect to multiple servers. Whenever your MCP host or client needs a tool, it connects to the MCP server. The MCP server then connects to, for example, a database or an API. MCP hosts and servers connect with each other through the MCP protocol.
+
+### Remote versus local MCP servers
+
+MCP utilizes a client-host-server architecture built on [JSON-RPC 2.0 for messaging](https://modelcontextprotocol.io/docs/concepts/architecture). Communication between clients and servers occurs over defined transport layers, and supports primarily two modes of operation:
+
+* **Remote MCP servers** - MCP clients connect to MCP servers over the internet, establishing a connection using HTTP and server-sent events (SSE), and authorizing the MCP client access to resources on the user's account using OAuth.
+
+* **Local MCP servers** MCP clients connect to MCP servers on the same machine, using standard input/output as a local transport method.
diff --git a/articles/api-center/includes/media/about-mcp-servers/mcp-architecture.png b/articles/api-center/includes/media/about-mcp-servers/mcp-architecture.png
diff --git a/articles/api-center/register-discover-mcp-server.md b/articles/api-center/register-discover-mcp-server.md
@@ -14,36 +14,7 @@ ms.author: danlep
 
 This article describes how to use Azure API Center to maintain an inventory (or *registry*) of remote model context protocol (MCP) servers and help stakeholders discover them using the API Center portal. MCP servers expose backend APIs or data sources in a standard way to AI agents and models that consume them.
 
-## About MCP servers
-
-AI agents are becoming widely adopted because of enhanced large language model (LLM) capabilities. However, even the most advanced models face limitations because of their isolation from external data. Each new data source potentially requires custom implementations to extract, prepare, and make data accessible for the models.
-
-The [model context protocol](https://www.anthropic.com/news/model-context-protocol) (MCP) helps solve this problem. MCP is an open standard for connecting AI models and agents with external data sources such as local data sources (databases or computer files) or remote services (systems available over the internet, such as remote databases or APIs).
-
-### MCP architecture
-
-The following diagram illustrates the MCP architecture:
- 
-:::image type="content" source="media/register-discover-mcp-server/mcp-architecture.png" alt-text="Diagram of model context protocol (MCP) architecture.":::
-
-The architecture consists of the following components:
-
-| Component      | Description                                                                                     |
-|----------------|-------------------------------------------------------------------------------------------------|
-| **MCP hosts**  | LLM applications such as chat apps or AI assistants in your IDEs (like GitHub Copilot in Visual Studio Code) that need to access external capabilities |
-| **MCP clients**| Protocol clients, inside the host application, that maintain 1:1 connections with servers        |
-| **MCP servers**| Lightweight programs that each expose specific capabilities and provide context, tools, and prompts to clients |
-| **MCP protocol**| Transport layer in the middle        |
-
-MCP follows a client-server architecture where a host application can connect to multiple servers. Whenever your MCP host or client needs a tool, it connects to the MCP server. The MCP server then connects to, for example, a database or an API. MCP hosts and servers connect with each other through the MCP protocol.
-
-### Remote versus local MCP servers
-
-MCP utilizes a client-host-server architecture built on [JSON-RPC 2.0 for messaging](https://modelcontextprotocol.io/docs/concepts/architecture). Communication between clients and servers occurs over defined transport layers, and supports primarily two modes of operation:
-
-* **Remote MCP servers** - MCP clients connect to MCP servers over the internet, establishing a connection using HTTP and server-sent events (SSE), and authorizing the MCP client access to resources on the user's account using OAuth.
-
-* **Local MCP servers** MCP clients connect to MCP servers on the same machine, using standard input/output as a local transport method.
+[!INCLUDE [about-mcp-servers](includes/about-mcp-servers.md)]
 
 ## MCP servers in your API inventory
 
diff --git a/articles/api-management/export-rest-mcp-server.md b/articles/api-management/export-rest-mcp-server.md
@@ -12,6 +12,12 @@ ms.author: danlep
 
 [!INCLUDE [api-management-availability-all-tiers](../../includes/api-management-availability-all-tiers.md)]
 
+<!-- Any limitations on SKUs
+
+Document need an IcM for SKUv2 --
+
+Need to use the GenAI release group -->
+
 In API Management, you can expose a REST API managed in API Management as a remote [Model Context Protocol (MCP)](https://www.anthropic.com/news/model-context-protocol) server. You can expose one or more of the API operations as tools that can be called by clients using the MCP protocol. 
 
 > [!NOTE]
