菜单
This skill should be used when the user asks to "list clusters", "show clusters", "cluster overview", "list projects", "which clusters", "compare clusters", "cluster status", "集群列表", "项目列表", "集群概览", "集群状态", "集群对比", or discusses Rancher cluster and project navigation (集群和项目管理).
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
This skill should be used when the user asks to "inspection", "inspect cluster", "health check", "patrol", "cluster review", "巡检", "集群巡检", "健康检查", "集群体检", "日常巡检", "安全巡检", "变更前检查", "变更后检查", "定期检查", or discusses systematic cluster health inspection and patrol (系统化集群巡检和健康检查).
This skill should be used when the user asks to "node capacity", "cluster capacity", "resource usage", "node health", "overloaded", "node status", "how much capacity", "resource utilization", "容量", "节点分析", "资源使用", "节点健康", "负载", "扩容", "容量规划", or discusses cluster capacity planning and node resource analysis (集群容量规划和节点资源分析).
This skill should be used when the user asks to "rollout history", "deployment changes", "watch deployment", "diff", "what changed", "compare deployments", "monitor rollout", "deployment status", "cross cluster diff", "部署历史", "发布历史", "变更", "部署对比", "监控部署", "跨集群对比", or discusses deployment tracking, change management, and release monitoring (部署追踪、变更管理和发布监控).
This skill should be used when the user asks to "list resources", "find resources", "get all", "dependency tree", "what depends on", "show dependencies", "resource inventory", "what's in namespace", "recently created", "资源列表", "依赖关系", "资源清单", "命名空间资源", "最近创建", or discusses Kubernetes resource discovery, inventory, and dependency analysis (资源发现、清单和依赖分析).
This skill should be used when the user asks to "diagnose pod", "pod logs", "why is pod failing", "check events", "troubleshoot", "investigate", "pod not ready", "crashloopbackoff", "pod restart", "container logs", "what's wrong with", "诊断 Pod", "排查问题", "查看日志", "为什么 Pod 失败", "事件查询", or discusses Kubernetes resource troubleshooting and diagnostics (资源排查和诊断).
| name | rancher-cluster-management |
| description | This skill should be used when the user asks to "list clusters", "show clusters", "cluster overview", "list projects", "which clusters", "compare clusters", "cluster status", "集群列表", "项目列表", "集群概览", "集群状态", "集群对比", or discusses Rancher cluster and project navigation (集群和项目管理). |
| version | 1.0.0 |
导航和管理 Rancher 多集群环境。简单查询直接执行；复杂的多集群分析委托给 Sub-Agent。
这些简单操作立即执行：
| 操作 | 工具 | 何时直接使用 |
|---|---|---|
| 列出集群 | mcp__rancher__cluster_list | 用户要求查看集群列表 |
| 按名称搜索集群 | mcp__rancher__cluster_list | 提供集群名称关键词 |
| 列出项目 | mcp__rancher__project_list | 用户要求查看项目列表 |
| 按集群列出项目 | mcp__rancher__project_list | 提供集群 ID |
rancher-cluster-explorer用于: 集群概览、容量分析、多集群对比
何时委托:
参数:
{
"action": "cluster_overview" | "compare_clusters",
"cluster": "c-abc123",
"clusters": ["c-abc123", "c-def456"]
}
用户请求：
├─ "列出集群" / "list clusters" / "show all clusters"
│ └─ 直接使用 cluster_list
│
├─ "列出项目" / "list projects" + 集群名/ID
│ └─ 直接使用 project_list
│
├─ "集群概览" / "cluster overview" / "cluster health"
│ └─ 委托给 rancher-cluster-explorer（包含容量和节点分析）
│
├─ "对比集群 A 和 B" / "compare clusters"
│ └─ 委托给 rancher-cluster-explorer（并行分析多集群）
│
└─ "哪个集群有空间" / "which cluster has capacity"
└─ 委托给 rancher-cluster-explorer（并行获取容量）
用户: "对比 production 和 staging 集群"
→ 并行启动：
Agent 1: rancher-cluster-explorer（集群 A 概览）
Agent 2: rancher-cluster-explorer（集群 B 概览）
→ 合并对比展示
用户: "所有集群的状态"
→ 步骤 1: 直接调用 cluster_list 获取所有集群
→ 步骤 2: 为每个集群并行启动 explorer Agent
→ 步骤 3: 汇总展示
如果用户提供集群名称而非 ID：
→ 使用 cluster_list（name: "关键词"）搜索
→ 获取匹配的集群 ID
## 可用集群
| 集群 | ID | 状态 | 节点数 | Provider |
|------|-----|------|--------|----------|
| production | c-abc123 | Active | 5 | Custom |
| staging | c-def456 | Active | 3 | Custom |
## 集群概览: production (c-abc123)
### 基本信息
- 状态: Active
- 节点数: 5
- 项目数: 3
### 资源容量
| 资源 | 请求 | 限制 | 实际使用 |
|------|------|------|----------|
| CPU | 65% | 120% | 45% |
| 内存 | 70% | 90% | 60% |
### 节点状态
| 节点 | 状态 | CPU | 内存 | Pod数 |
|------|------|-----|------|-------|
| node-1 | Ready | 55% | 62% | 42 |
| node-2 | Ready | 72% | 78% | 38 |