docs: update install docs and ArtifactHub metadata to headlamp namespace

- README: update Helm install URL to v1.0.0, add pods/proxy RBAC scope notes
- docs/getting-started/installation.md: update all download URLs from v0.2.4 to v1.0.0, remove obsolete sha256sum example, fix build output placeholder
- docs/getting-started/quick-start.md: update Helm URL to v1.0.0
- docs/deployment/helm.md: update install command and values URLs to v1.0.0
- docs/README.md: fix Artifact Hub URL (was duplicated 'headlamp')
- docs/architecture/overview.md: update namespace references from kube-system to headlamp
- CLAUDE.md: update driver namespace constant from kube-system to headlamp
- CHANGELOG.md: add [Unreleased] entry documenting the namespace reference update

Co-Authored-By: Paperclip <noreply@paperclip.ing>

README.md

@@ -63,12 +63,12 @@ config:
 pluginsManager:
   sources:
     - name: tns-csi
-      url: https://github.com/privilegedescalation/headlamp-tns-csi-plugin/releases/download/v0.2.4/tns-csi-0.2.4.tar.gz
+      url: https://github.com/privilegedescalation/headlamp-tns-csi-plugin/releases/download/v1.0.0/tns-csi-1.0.0.tar.gz
 ```
 
 ## RBAC / Security Setup
 
-The plugin reads from the Kubernetes API and the tns-csi controller pod's Prometheus endpoint. The Benchmark page additionally creates and deletes Jobs and PVCs.
+The plugin reads from the Kubernetes API and the tns-csi controller pod's Prometheus endpoint (deployed in `kube-system`). The Benchmark page additionally creates and deletes Jobs and PVCs.
 
 ### Minimal read-only permissions
 
@@ -90,6 +90,10 @@ rules:
   - apiGroups: [""]
     resources: ["pods/log"]
     verbs: ["get"]
+  - apiGroups: [""]
+    resources: ["pods/proxy"]
+    verbs: ["get"]
+    resourceNames: ["pods"]
 ```
 
 ### Additional permissions for Benchmark page
@@ -105,13 +109,13 @@ rules:
 
 ### Metrics access
 
-The plugin fetches Prometheus metrics from the tns-csi controller pod via the Kubernetes pod proxy sub-resource. Grant `get` on `pods/proxy` in `kube-system`:
+The plugin fetches Prometheus metrics from the tns-csi controller pod via the Kubernetes pod proxy sub-resource in `kube-system`. Grant `get` on `pods/proxy` scoped to `kube-system`:
 
 ```yaml
   - apiGroups: [""]
     resources: ["pods/proxy"]
     verbs: ["get"]
-    # Optionally scope to the controller pod namespace
+    # Scope to kube-system where the tns-csi controller runs
 ```
 
 Apply the role and bind it to your Headlamp service account with a ClusterRoleBinding.