skyhook-io / radar
A single-binary Kubernetes UI with built-in MCP server that lets AI agents inspect cluster topology, events, and configuration.
星标趋势
AI 分析
项目摘要
Radar is an open-source Kubernetes UI delivered as a single Go binary that provides topology visualization, event timelines, Helm/GitOps management, traffic monitoring, and cluster audits. Its distinguishing feature is a built-in MCP server that exposes cluster context to AI agents, enabling AI-driven inspection and troubleshooting.
为什么值得关注
It uniquely combines a comprehensive Kubernetes dashboard with native AI agent integration via MCP, allowing both humans and AI systems to inspect cluster state, topology, and events through a single binary that requires zero cluster-side installation and keeps all data local for privacy.
优势
- Single binary with zero cluster-side installation and no dependencies
- Built-in MCP server enabling AI agent access to cluster context
- Comprehensive views including topology, timeline, GitOps, traffic, and audit
- Privacy-first design keeping all cluster data on local machine
- Active development with 63 releases in 6 months and 54 contributors
局限性
- No automated tests despite active development
- No examples provided for MCP server integration or advanced workflows
- Relatively new project with limited production track record
使用场景
- Kubernetes cluster troubleshooting and root cause analysis
- AI agent-driven cluster inspection and automated diagnostics
- GitOps workflow visualization for ArgoCD and Flux
- Cluster security auditing and RBAC review
- Helm release comparison and upgrade impact analysis