Ocp admin/multi cluster report

.claude-plugin/marketplace.json

@@ -45,6 +45,18 @@
       "source": "./rh-virt",
       "category": "virtualization",
       "skills": "./skills"
+    },
+    {
+      "name": "ocp-admin",
+      "description": "Red Hat OpenShift Administration Agentic Skills Collection",
+      "version": "1.0.0",
+      "author": {
+        "name": "Red Hat Ecosystem Engineering",
+        "email": "eco-engineering@redhat.com"
+      },
+      "source": "./ocp-admin",
+      "category": "openshift",
+      "skills": "./skills"
     }
   ]
 }

README.md

@@ -174,6 +174,12 @@ Add the server configuration to `<pack>/.mcp.json`:
 - ❌ Never hardcode API keys, tokens, or secrets
 - ✅ Set appropriate security isolation level
 
+**Platform Notes (Linux vs macOS):**
+
+On **Linux**, you may want to add `--userns=keep-id:uid=65532,gid=65532` to the Podman `args` for proper UID/GID mapping inside the container. This ensures the container process runs with the correct non-root user identity.
+
+On **macOS**, this flag is **not supported** because Podman runs inside a Linux VM where user namespace mapping behaves differently. Omit it on macOS or the container will fail to start.
+
 ### Step 2: Add Custom Metadata (Optional)
 
 To display repository links and tool descriptions on the documentation site, add an entry to `docs/mcp.json`:

ocp-admin/.mcp.json

@@ -7,7 +7,6 @@
         "--rm",
         "-i",
         "--network=host",
-        "--userns=keep-id:uid=65532,gid=65532",
         "-v", "${KUBECONFIG}:/kubeconfig:ro,Z",
         "--entrypoint", "/app/kubernetes-mcp-server",
         "quay.io/ecosystem-appeng/openshift-mcp-server:latest",

ocp-admin/README.md

@@ -16,6 +16,17 @@ Administration and management tools for OpenShift Container Platform. This pack
 - OpenShift cluster access via `KUBECONFIG`
 - For multi-cluster reports, a kubeconfig with multiple contexts
 
+## Helper Scripts
+
+The `cluster-report` skill uses two Python scripts (stdlib only, no dependencies) in `scripts/cluster-report/`:
+
+| Script | Purpose |
+|--------|---------|
+| `assemble.py` | Resolves `$file` references in the manifest JSON, loading persisted MCP output from disk into a complete data structure. With `--aggregate`, pipes into `aggregate.py` automatically. |
+| `aggregate.py` | Computes per-cluster and fleet-wide metrics (CPU/memory usage, pod status counts, GPU totals, top namespaces) and flags attention items (>85% utilization, failed pods, missing metrics). |
+
+Both scripts read from stdin and write to stdout. They are invoked as a pipeline by the skill and should be treated as black boxes.
+
 ## MCP Servers
 
 - **openshift** - OpenShift cluster management with multi-cluster support