Kubernetes FAQs

Question 1

What is `mcp-kubernetes-ro`?

Accepted Answer

`mcp-kubernetes-ro` is a Model Context Protocol (MCP) server that provides secure, read-only access to Kubernetes clusters. It's specifically designed to integrate seamlessly with AI assistants, allowing them to query and understand your cluster's state without risk of destructive operations.

Question 2

Does `mcp-kubernetes-ro` require the `kubectl` binary?

Accepted Answer

No, a key feature of `mcp-kubernetes-ro` is that it does not require the `kubectl` binary to be installed. It leverages your local `kubectl` configuration file (`~/.kube/config`) to connect to your clusters, offering flexibility and ease of use.

Question 3

What kind of information can I access with this tool?

Accepted Answer

You can access a wide range of Kubernetes information including listing any resource types, retrieving detailed resource configurations, accessing pod logs with advanced filtering, discovering API resources, and even getting node/pod metrics (requires Metrics Server in your cluster). It also offers base64 encoding/decoding utilities.

Question 4

How does `mcp-kubernetes-ro` ensure security?

Accepted Answer

Security is paramount. The tool enforces strict read-only access to your Kubernetes clusters, preventing any destructive operations or modifications. This design allows for comprehensive cluster inspection and troubleshooting capabilities while maintaining a secure environment for AI interactions.

Question 5

How can I integrate `mcp-kubernetes-ro` with an AI assistant or editor?

Accepted Answer

`mcp-kubernetes-ro` supports multiple transport modes, including standard I/O (stdio) for editor integrations and Server-Sent Events (SSE) for web-based applications. You can configure it directly within compatible editor settings or run it as an HTTP server to expose its capabilities to AI models.

Kubernetes

关于

主要功能

使用案例