MCP.so
登录

Kubectl MCP Tool

@rohitg00

关于 Kubectl MCP Tool

Published in CNCF Landscape: A MCP server for Kubernetes.

基本信息

分类

其他

许可证

MIT

运行时

python

传输方式

stdio

发布者

rohitg00

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "kubectl-mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "kubectl-mcp-server"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Kubectl MCP Tool?

Kubectl MCP Tool lets you control Kubernetes infrastructure through natural language conversations with AI. It integrates with any MCP-compatible AI assistant and kubectl to manage clusters, debug issues, deploy apps, and audit security.

How to use Kubectl MCP Tool?

Install via npx (npx -y kubectl-mcp-server) or pip (pip install kubectl-mcp-server), then add the server configuration to your AI assistant’s MCP settings file (e.g., claude_desktop_config.json or .cursor/mcp.json). Once connected, use natural language commands to interact with your cluster.

Key features of Kubectl MCP Tool?

  • 253 powerful tools and 8 workflow prompts
  • AI-powered crash and network diagnostics
  • Built-in cost optimization analysis
  • Enterprise OAuth 2.1 authentication and RBAC validation
  • Interactive HTML dashboards with live metrics
  • Works with all major MCP-compatible AI assistants

Use cases of Kubectl MCP Tool?

  • Debugging a crashing pod with logs, events, and resource analysis
  • Deploying a Redis cluster with best practices via natural language
  • Auditing RBAC permissions and security across all namespaces
  • Identifying idle resources and cost optimization opportunities
  • Visualizing cluster dashboards with live metrics and event timelines

FAQ from Kubectl MCP Tool

What are the dependencies to run Kubectl MCP Tool?

You need Python 3.9+ (for pip) or Node.js 14+ (for npx), kubectl installed and configured, and access to a Kubernetes cluster.

Which AI assistants are supported?

Kubectl MCP Tool works natively with Claude Desktop, Cursor, Windsurf, GitHub Copilot, Gemini CLI, Goose, Roo Code, and 15+ other MCP-compatible clients.

Can I manage multiple Kubernetes clusters?

Yes, the

评论

其他 分类下的更多 MCP 服务器