Radar Kubernetes MCP Server — Claude Code MCP server: features
Connects an AI agent to Kubernetes for cluster inspection, troubleshooting, and operations through a built-in MCP server.
Open on GitHubWhat Radar Kubernetes MCP Server does
Radar gives AI agents access to Kubernetes through its built-in MCP server. An agent can inspect cluster resources, investigate problems, and carry out operations. The project also describes GitOps support and RBAC-enforced access.
Radar runs as a single binary on your machine and talks directly to the Kubernetes API. Cluster data stays on that machine. You’ll need Radar installed and access to the Kubernetes API; the quick install is curl -fsSL https://get.radarhq.io | sh, followed by kubectl radar. The installation instructions don’t include a step for connecting the MCP server to Claude Code, so these commands alone don’t establish that connection.
What it helps with
- inspect Kubernetes resources
- investigate a cluster issue
- run an operation with RBAC controls
- check GitOps changes
Example prompts for the agent
- Inspect the resources in my Kubernetes cluster.
- Investigate what caused this cluster issue.
- Run this operation using the available RBAC permissions.
How to install Radar Kubernetes MCP Server
Install
The MCP registry has no launch command for it: see the repository README
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Radar Kubernetes MCP Server
How do I install Radar Kubernetes MCP Server in Claude Code?
Install Radar with curl -fsSL https://get.radarhq.io | sh, then run kubectl radar. These steps install and launch Radar, but don’t describe connecting its MCP server to Claude Code.
What does Radar Kubernetes MCP Server do?
It lets AI agents inspect a Kubernetes cluster and perform operations through a built-in MCP server. The project also describes GitOps support and RBAC-enforced access.
Where does Radar run, and where does cluster data stay?
Radar runs on your machine and connects directly to the Kubernetes API. Cluster data stays on that machine.