Radar is a single-binary, open-source Kubernetes UI with an MCP server for AI agents, offering topology, resources, timeline, Helm, GitOps, traffic, audits, and RBAC. It runs locally from a laptop or in-cluster, with no account or cloud dependency.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
Radar is described as a missing open-source Kubernetes UI in a single binary with an integrated MCP server for AI agents. It provides topology, resources, timeline, Helm management, GitOps, traffic, audit features, and MCP context for AI agents. It is designed to run on a laptop or in-cluster without requiring an account or cloud connectivity.
How it works
Radar exposes views for topology, resources, image filesystem, timeline, Helm, GitOps, traffic, cost, audit, upgrade impact, RBAC, and MCP context. It communicates with the Kubernetes API directly from the client side (local-first). It can be installed via curl pipe to install the binary and then run kubectl radar, or installed via various package managers or in-cluster Helm deployment. The MCP server is built-in to support AI tool integration.
Getting started
Install and run in 30 seconds:
curl -fsSL https://get.radarhq.io | sh && kubectl radar
Quick Install:
curl -fsSL https://get.radarhq.io | sh
Homebrew:
brew install skyhook-io/tap/radar
In-Cluster Deployment (Helm):
helm repo add skyhook https://skyhook-io.github.io/helm-charts
helm install radar skyhook/radar -n radar --create-namespace
Usage examples show how to inspect in-cluster Radar Cloud installations and how to run radar commands locally. The README lists CLI flags for configuration, including kubeconfig paths, namespace filters, server port, timeline storage, MCP toggles, authentication, and more.
Recent releases
Latest release: v1.9.0 on 2026-08-05 with Radar v1.9.0, described as modern Kubernetes visibility, local-first, and fast. Prior releases v1.8.7, v1.8.6, v1.8.5, v1.8.4 share the same feature summary (Topology, event timeline, service traffic, resource browsing, Helm management).
Traction
Stars: 2723
Behind the repo
Not provided in the README snapshot.
Caveats
License: Apache-2.0 Created: 2026-01-20. Last push: 2026-08-04. Open issues: 49. Language: Go. Topics include ai-agents, argocd, gitops, helm, k8s, mcp, MCP server. In-cluster deployment supported via Helm with Gateway API HTTPRoute options. The README includes detailed configuration options for performance tuning and environment overrides.