feat: add vm-lifecycle-manager ASE evaluation submission

submissions/vm-lifecycle-manager/evals/evals.json

@@ -0,0 +1,17 @@
+{
+  "skill_name": "vm-lifecycle-manager",
+  "evals": [
+    {
+      "id": "vm-lifecycle-manager-eval",
+      "name": "VM Lifecycle Management",
+      "prompt": "You are an OpenShift Virtualization administrator. A developer asks you to restart VM `api-server` in namespace `prod`.\n\nDescribe your methodology for handling VM lifecycle operations (start, stop, restart). What confirmation steps do you require? How do you implement restart specifically, and why? What error scenarios do you handle?",
+      "expected_output": "A lifecycle management approach that requires explicit user confirmation before any action, implements restart as two separate stop-then-start operations to avoid resourceVersion conflicts, and handles edge cases like VMs already in the desired state.",
+      "assertions": [
+        "The output requires explicit user confirmation (yes/no) before executing any lifecycle operation, presenting the operation details and impact first.",
+        "The output implements restart as two separate sequential operations (stop then start) rather than a single restart command, and explains this avoids resourceVersion conflicts.",
+        "The output describes verifying the VM has fully stopped before issuing the start command during a restart sequence.",
+        "The output describes how to handle the case where a VM is already in the desired state (e.g., already running when start is requested)."
+      ]
+    }
+  ]
+}

submissions/vm-lifecycle-manager/metadata.yaml

@@ -0,0 +1,5 @@
+name: vm-lifecycle-manager
+description: "Manage virtual machine lifecycle operations including start, stop, and"
+persona: rh-virt
+version: "1.0.0"
+eval_engine: ase

submissions/vm-lifecycle-manager/skills/SKILL.md

@@ -0,0 +1,314 @@
+---
+name: vm-lifecycle-manager
+description: 'Manage virtual machine lifecycle operations including start, stop, and
+  restart.
+
+
+  Use when:
+
+  - "Start VM [name]"
+
+  - "Stop the virtual machine [name]"
+
+  - "Restart VM [name]"
+
+  - "Power on/off VM [name]"
+
+
+  This skill handles VM state transitions safely with user confirmation for each action.
+
+
+  NOT for creating VMs (use vm-create) or deleting VMs (use vm-delete).
+
+  '
+license: Apache-2.0
+---
+
+# /vm-lifecycle-manager Skill
+
+Control virtual machine power state in OpenShift Virtualization using the `vm_lifecycle` tool.
+
+## Prerequisites
+
+**Required MCP Server**: `openshift-virtualization` ([OpenShift MCP Server](https://github.com/openshift/openshift-mcp-server))
+
+**Required MCP Tools**:
+- `vm_lifecycle` (from openshift-virtualization) - Manage VM power state
+
+**Required Environment Variables**:
+- `KUBECONFIG` - Path to Kubernetes configuration file with cluster access
+
+**Required Cluster Setup**:
+- OpenShift cluster (>= 4.19)
+- OpenShift Virtualization operator installed
+- ServiceAccount with RBAC permissions to update VirtualMachine resources
+
+### Prerequisite Verification
+
+**Before executing:**
+1. Check `openshift-virtualization` exists in `mcps.json` → If missing, report setup
+2. Verify `KUBECONFIG` is set (presence only, never expose value) → If missing, report
+
+**Human Notification Protocol:** `❌ Cannot execute vm-lifecycle-manager: MCP server not available. Setup: Add to mcps.json, set KUBECONFIG, restart Claude Code. Docs: https://github.com/openshift/openshift-mcp-server`
+
+⚠️ **SECURITY**: Never display KUBECONFIG path or credential values.
+
+## When to Use This Skill
+
+**Trigger when:**
+- User explicitly invokes `/vm-lifecycle-manager` command
+- User requests starting/stopping/restarting a VM
+- User wants to change VM power state
+
+**User phrases:**
+- "Start VM web-server in namespace vms"
+- "Stop the database VM"
+- "Restart test-vm"
+- "Power on the VM called api-server"
+- "/vm-lifecycle-manager" (explicit command)
+
+**Do NOT use when:**
+- Create VM → `/vm-create`
+- List VMs → `/vm-inventory`
+- Delete VM → `/vm-delete`
+
+## Workflow
+
+### Step 1: Gather Parameters and Confirm
+
+**Required from user:** VM Name, Namespace, Action (start|stop|restart)
+
+**Present for confirmation:**
+
+```markdown
+## VM Lifecycle Operation
+
+| Parameter | Value | Impact |
+|-----------|-------|--------|
+| VM Name | `<vm>` | from user |
+| Namespace | `<ns>` | from user |
+| Action | `<action>` | start: consumes resources / stop: graceful shutdown / restart: brief interruption (~1-2min) |
+
+Confirm: yes/no
+```
+
+**WAIT** for explicit "yes" before proceeding to Step 2.
+
+### Step 2: Execute Lifecycle Operation
+
+**ONLY AFTER user confirmation in Step 1.**
+
+**For start or stop actions:**
+
+**MCP Tool**: `vm_lifecycle` (namespace=`<ns>`, name=`<vm>`, action=`<start|stop>`)
+
+**For restart action (composite operation):**
+
+**CRITICAL**: Implement restart as two separate operations to avoid resourceVersion conflicts:
+
+1. **Stop VM**: `vm_lifecycle` (namespace=`<ns>`, name=`<vm>`, action="stop")
+2. **Verify stopped**: `resources_get` (apiVersion="kubevirt.io/v1", kind="VirtualMachine", namespace=`<ns>`, name=`<vm>`) → Check `status.printableStatus` == "Stopped"
+3. **Wait**: 5 seconds for VM to fully stop
+4. **Start VM**: `vm_lifecycle` (namespace=`<ns>`, name=`<vm>`, action="start")
+5. **Verify started**: `resources_get` → Check `status.printableStatus` == "Running"
+
+**Errors:**
+- VM not found → Report, suggest vm-inventory
+- Permission denied → Report RBAC error
+- Already in desired state → Inform user
+- Stop fails during restart → Report, do not proceed to start
+- Start fails during restart → Report, VM is stopped
+- Transition fails → Report details
+
+### Step 3: Report Operation Status
+
+**On Success:**
+
+```markdown
+## ✓ VM <Action> Successful
+
+**VM**: `<vm>` | **Namespace**: `<ns>` | **Action**: <action> | **RunStrategy**: <Always|Halted>
+
+**Impact**:
+- **start**: Running, consuming resources (CPU/memory). Access: virtctl console or SSH. RunStrategy: Always (auto-restart on crash)
+- **stop**: Stopped, resources freed. State preserved. Start: "Start VM <vm>". RunStrategy: Halted (stays off)
+- **restart**: Running after stop+start. Brief interruption (~1-2min). Monitor app logs. RunStrategy: Always
+
+**Next**: "Show status of VM <vm>" or "List VMs in namespace <ns>"
+```
+
+**On Failure:**
+
+**OPTIONAL**: Read [lifecycle-errors.md](../../docs/troubleshooting/lifecycle-errors.md) for start/stop failures or [scheduling-errors.md](../../docs/troubleshooting/scheduling-errors.md) for ErrorUnschedulable. Output: "Consulted lifecycle-errors.md for failure."
+
+**When to consult**: Start/stop failures, stuck transitions, unexpected errors. **NOT**: Already in state, not found, RBAC errors.
+
+```markdown
+## ❌ Lifecycle Operation Failed
+
+**Error**: <error>
+
+**Causes**: VM not found | RBAC denied | Already in desired state | VM in transition (wait 30-60s) | Resource constraints (start)
+
+**Troubleshoot**:
+1. vm-inventory to verify VM exists
+2. Check RBAC: `oc auth can-i update virtualmachines -n <ns>`
+3. View VM status and events
+4. Check node capacity (for start operations)
+```
+
+## Common Issues
+
+### Issue 1: VM Not Found
+**Error**: "VirtualMachine 'xyz' not found in namespace 'abc'"
+**Solution**: Verify spelling, check namespace, use vm-inventory, VM may be deleted
+
+### Issue 2: VM Already in Desired State
+**Warning**: "VM is already running" (when attempting start)
+**Solution**: Not an error - VM already in desired state. Use `restart` if intended to restart
+
+### Issue 3: Permission Denied
+**Error**: "Forbidden: User cannot update VirtualMachines"
+**Solution**: Verify RBAC permissions (update VirtualMachine resources), contact admin
+
+### Issue 4: VM Stuck in Transitioning State
+**Error**: "VM stuck in 'Terminating' or 'Starting'"
+**Solution**: Wait 30-60s, check events (`oc describe vm`), use vm-troubleshooter, check virt-launcher pod
+
+### Issue 5: Insufficient Resources (Start)
+**Error**: "Insufficient CPU/memory to start VM"
+**Solution**: Check cluster availability, stop other VMs, scale nodes, resize VM to smaller instance type
+
+### Issue 6: Restart Implementation
+**Note**: Restart is implemented as two separate operations (stop → verify → start → verify)
+**Reason**: Avoids Kubernetes resourceVersion conflicts when using single restart action
+**Behavior**: If stop succeeds but start fails, VM remains stopped. Check VM status with vm-inventory
+
+## Understanding RunStrategy
+
+| Action | RunStrategy | Behavior |
+|--------|------------|----------|
+| start | Always | Runs, auto-restarts on crash |
+| stop | Halted | Stops, stays off |
+| restart | Always | Stops, starts, auto-restarts |
+
+## Dependencies
+
+### Required MCP Servers
+- `openshift-virtualization` - OpenShift MCP server with KubeVirt toolset
+
+### Required MCP Tools
+- `vm_lifecycle` - Manage VM power state (start/stop/restart)
+
+### Related Skills
+- `vm-create` - Create VMs
+- `vm-inventory` - Check VM status
+- `vm-troubleshooter` (planned) - Diagnose startup/shutdown issues
+
+### Reference Documentation
+- [lifecycle-errors.md](../../docs/troubleshooting/lifecycle-errors.md) - Start/stop failures, stuck transitions (consulted on failures)
+- [scheduling-errors.md](../../docs/troubleshooting/scheduling-errors.md) - ErrorUnschedulable, resource constraints (consulted when VM won't start)
+- [Troubleshooting INDEX](../../docs/troubleshooting/INDEX.md) - Navigation hub for error categories
+- [OpenShift Virt Docs](https://docs.redhat.com/en/documentation/openshift_container_platform/4.21/html-single/virtualization/index#virt/about_virt/about-virt.html)
+- [KubeVirt Lifecycle](https://kubevirt.io/user-guide/virtual_machines/lifecycle/)
+- [RunStrategy Docs](https://kubevirt.io/user-guide/virtual_machines/run_strategies/)
+- [OpenShift MCP](https://github.com/openshift/openshift-mcp-server)
+
+## Critical: Human-in-the-Loop Requirements
+
+**IMPORTANT:** This skill requires explicit user confirmation before executing. You MUST:
+
+1. **Wait for user confirmation** on specific action (start/stop/restart) before executing `vm_lifecycle`
+2. **Do NOT proceed** until user explicitly approves
+3. **Present action clearly** with VM name, namespace, intended state change
+4. **Never auto-execute** - these operations impact running services
+5. **Warn about consequences** - stopping interrupts services, starting consumes resources, restarting causes downtime
+
+If user says "no" or wants to reconsider, do not proceed.
+
+**Why**: start (consumes resources), stop (interrupts services), restart (brief downtime). User should verify correct VM and understand impact.
+
+## Security Considerations
+
+- Requires RBAC permissions (update VirtualMachine resources)
+- Stop operations attempt graceful shutdown; guest OS/app data consistency depends on workload
+- Start operations consume resources (quota limits apply)
+- All operations audited in Kubernetes API logs
+- KUBECONFIG credentials never exposed
+- ServiceAccount permissions limit controllable VMs
+
+## Example Usage
+
+### Example 1: Start Stopped VM
+
+```
+User: "Start VM web-server in namespace vms"
+Agent: [vm-lifecycle-manager skill]
+## VM Lifecycle Operation
+| VM Name | `web-server` | Namespace | `vms` | Action | `start` | consumes resources |
+Confirm: yes/no
+User: "yes"
+Agent: [vm_lifecycle(namespace="vms", name="web-server", action="start")]
+## ✓ VM Start Successful
+VM: `web-server` | Namespace: `vms` | RunStrategy: Always
+Impact: Running, consuming resources. Access: virtctl console or SSH
+```
+
+### Example 2: Stop with User Reconsidering
+
+```
+User: "Stop the database VM in production namespace"
+Agent: [vm-lifecycle-manager skill]
+| VM | `database` | Namespace | `production` | Action | `stop` | graceful shutdown |
+Confirm: yes/no
+User: "wait, no, don't stop it"
+Agent: Operation cancelled. VM 'database' in namespace 'production' remains running.
+```
+
+### Example 3: Restart VM
+
+```
+User: "Restart test-vm in namespace dev"
+Agent: [vm-lifecycle-manager skill]
+| VM | `test-vm` | Namespace | `dev` | Action | `restart` | brief interruption ~1-2min |
+User: "yes"
+Agent: [Composite restart operation]
+       1. vm_lifecycle(namespace="dev", name="test-vm", action="stop")
+       2. Verify stopped: resources_get → status.printableStatus="Stopped"
+       3. Wait 5s
+       4. vm_lifecycle(namespace="dev", name="test-vm", action="start")
+       5. Verify started: resources_get → status.printableStatus="Running"
+## ✓ VM Restart Successful
+VM: `test-vm` | Namespace: `dev` | RunStrategy: Always
+Impact: Running after stop+start. Brief interruption. Monitor app logs.
+```
+
+### Example 4: VM Already in Desired State
+
+```
+User: "Start web-server in namespace vms"
+Agent: [vm-lifecycle-manager skill]
+       [vm_lifecycle(action="start")]
+## ℹ️ VM Already Running
+VM: `web-server` | Namespace: `vms` | Status: Running
+Result: No action taken - VM already in desired state.
+To restart: "Restart VM web-server in namespace vms"
+```
+
+### Example 5: Batch Operations
+
+```
+User: "Stop VMs web-01, web-02, web-03 in namespace production"
+Agent: [vm-lifecycle-manager skill - batch mode]
+## Batch Lifecycle Operation
+Stopping 3 VMs in 'production': web-01, web-02, web-03
+Impact: All 3 VMs will shut down, services interrupted.
+Confirm: yes/no
+User: "yes"
+Agent: [Executes vm_lifecycle for each VM sequentially]
+## ✓ Batch Stop Successful
+- web-01: Stopped
+- web-02: Stopped
+- web-03: Stopped
+All VMs stopped. Resources freed.
+```