Paul Yu

It's been a few months since the AKS-MCP server was announced. Since then, there have been several updates and improvements. The MCP server can be easily installed on a local machine using the AKS Extension for VS Code, or via the GitHub MCP registry, or even using the Docker MCP hub.

In this blog post, I'll show you one approach to running the AKS MCP server: deploying it inside an AKS cluster as a Streamable HTTP service. This pattern demonstrates how MCP servers can be centrally managed and made accessible to multiple clients—including AI assistants, automation tools, and even autonomous agents.

Kubernetes is great at orchestrating containers, but it can be a bit tricky to manage traffic routing. There are many options and implementations that you, as a cluster operator have probably had to deal with. We have the default Service resource that can be used to expose applications, but it is limited to routing based on layer 4 (TCP/UDP) and does not support more advanced traffic routing use cases. There's also the Ingress controller, which enabled layer 7 (HTTP) routing, and securing the North-South traffic using TLS, but it was not standardized and each vendor implementation required learning a new set of resource annotations. When it comes to managing and securing East-West traffic between services, there's Service Mesh which is yet another layer of infrastructure to manage on top of Kubernetes. And we're in the same boat when it comes to resource management with each vendor having their own ways of doing things.