Manage multiple Kubernetes clusters through MCP tools for resources, workloads, nodes, pods, logs, and Prometheus queries.
Copy the AI prompt to install this server into Claude Code, Cursor, or another agent — or use 1-click editor setup below.
💡 Paste the JSON block into your client's configuration file under mcpServers, then restart the application.
Inspect callable tools, capabilities, and parameters exposed to AI agents by Kom.
The weibaohui/kom MCP server connects MCP clients to Kubernetes management functions across registered clusters. Its documented tool set covers cluster listing, dynamic resources, Deployments, DaemonSets, Nodes, Events, IngressClasses, and Pods. Resource operations include listing, retrieval, description, deletion, annotation, labeling, and patching, with support for custom resource definitions alongside built-in Kubernetes types.
Workload and node operations go beyond basic CRUD. Deployment tools can scale, restart, stop, restore, pause, resume, and roll back rollouts, update image tags, inspect rollout status and history, and retrieve related events or HPA data. Node tools cover taints, scheduling state, draining, resource usage, IP usage, top-node results, and pod counts. Pod tools can execute commands, inspect events and linked services or ingress objects, list or delete files, and read runtime environment information.
The library also documents SQL-style Kubernetes resource queries, cross-namespace queries, optional query caching, and Prometheus queries against an in-cluster or external Prometheus service. These capabilities are relevant when an agent needs operational inspection rather than only direct API calls.
Applications register Kubernetes clusters through the Go API. The README shows in-cluster registration and registration from a kubeconfig path with a caller-selected cluster ID. After initialization, the MCP server can be started from Go with mcp.RunMCPServer, using a server name, version, and port.
The weibaohui/kom MCP server supports two transport modes. A compiled local executable can be used through stdio, while the SSE mode exposes an endpoint at /sse; the example uses port 9096. Multiple clusters are registered within the same server, and tools accept cluster, namespace, resource type, and object identifiers where applicable.
The project is written in Go and the README demonstrates importing github.com/weibaohui/kom and its callbacks package. The initialization callback is registered before clusters are configured. For external cluster access, KUBECONFIG can point to a kubeconfig file. If that variable is absent, the shown example falls back to the user’s .kube/config path; in-cluster registration is also demonstrated.
The documented source workflow builds an executable with go build main.go, then runs the resulting binary. MCP clients can connect by supplying the executable as a stdio command, or by configuring the SSE URL. The README specifically lists Cursor, Claude Desktop, and Windsurf as suitable MCP integrations, noting that Claude Desktop uses stdio in the described setup.
No standalone package-manager installation command for the MCP server is specified in the provided material. Deployment therefore requires building or embedding the Go project rather than assuming a published executable package.
The weibaohui/kom MCP server performs Kubernetes operations using the credentials available to its process, so the reachable clusters and allowed actions depend on kubeconfig or in-cluster authorization. The provided material does not define a separate MCP API-key or OAuth flow. It also does not provide a prebuilt package-manager command or a hosted SSE service; the SSE address shown is produced by a locally built and started application.
The README contains a tool table documenting 59 MCP tools, while the shorter project description refers to nearly 50 built-in tools. The detailed table is the more specific count available in the supplied material.
Factual signals from GitHub, npm, and our automated checks — not a rating.
No reviews yet — be the first to share how this listing worked for you.
Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.
[](https://allmcps.com/mcp/weibaohui-kom)<a href="https://allmcps.com/mcp/weibaohui-kom"><img src="https://allmcps.com/api/badge/weibaohui-kom?style=directory" alt="Kom on AllMCPs" /></a>