skyhook-io/radar

What it does

Skyhook Radar is a Model Context Protocol (MCP) server that integrates with AI agents to provide comprehensive Kubernetes cluster visibility and management capabilities. It allows AI assistants to query and analyze cluster topology, resource utilization, events, logs, and Helm/GitOps deployment state across multiple Kubernetes clusters in real-time.

How to set up

Radar is deployed as a single binary with no additional agents or cloud dependencies required. Installation involves downloading the binary, configuring access to your Kubernetes clusters, and connecting it to your MCP-compatible AI tool. Configuration is straightforward with support for multiple cluster contexts and authentication methods.

Key features

• Query cluster topology and resource allocation across multiple clusters
• Real-time access to cluster events and logs
• Helm and GitOps state visibility
• Single binary deployment with no external agents
• No cloud dependency - runs locally
• Support for multiple Kubernetes clusters
• MCP-compatible for AI agent integration
• Apache 2.0 open-source license