Kubernetes MCP Server
About
A popular community Kubernetes MCP server that lets AI agents operate clusters in natural language — view/create/delete Pods, Deployments and Services, read logs, run kubectl and Helm operations, with read-only and non-destructive modes to protect production.
Our Verdict
Kubernetes MCP Server brings devops capabilities to the MCP ecosystem. Its GitHub popularity reflects community trust, and the standardized protocol ensures broad compatibility across AI clients.
Pricing
Use Cases
- Integrating devops capabilities into AI agent workflows
- Building custom AI-powered tools with Kubernetes MCP Server
- Enhancing AI agent functionality with standardized MCP access
Pros
- Kubernetes MCP Server provides standardized MCP protocol access
- Open source with 1,515 GitHub stars
- Active development and community support
Cons
- Requires MCP-compatible client (Claude, Cursor, etc.)
- Setup complexity varies by client
- Some features may depend on client capabilities
Installation
npx -y mcp-server-kubernetesFrequently Asked Questions
Which cluster does Kubernetes MCP connect to?
It reuses your local kubeconfig (by default ~/.kube/config and the current context), so it connects to whatever cluster kubectl currently points to. Switch context to switch clusters; permissions are entirely determined by that context's credentials, with no extra privilege escalation.
Could it accidentally delete production resources?
For that it offers read-only and non-destructive modes: when enabled, dangerous operations like delete are disabled and the agent can only view and make safe changes. For production, enable read-only or non-destructive mode and use a permission-limited context to keep risk contained.
Does it support Helm and many resource types?
Yes. Beyond CRUD on common resources like Pods/Deployments/Services/ConfigMaps, it can run general kubectl commands, view logs and events, and install/upgrade Helm charts, covering everyday troubleshooting and release operations.