diff --git a/scripts/create-cluster.sh b/scripts/create-cluster.sh
index 1035edf..eb7dd78 100755
--- a/scripts/create-cluster.sh
+++ b/scripts/create-cluster.sh
@@ -31,8 +31,10 @@ NETWORK_NAME="nebari-${CLUSTER_NAME}-net"
 GITOPS_DIR="/tmp/nebari-gitops-${CLUSTER_NAME}"
 
 if [[ "${PROFILE}" == "platform" ]]; then
-  # NIC's local provider uses MetalLB with a hardcoded 192.168.1.100-110 pool.
-  # Create a Docker network matching that subnet so MetalLB IPs are routable.
+  # Create a Docker network matching the MetalLB address pool configured in
+  # deploy-platform.sh (192.168.1.100-110) so MetalLB IPs are routable from
+  # the runner. The subnet and pool are kept in sync between this file and the
+  # NIC config (cluster.local.metallb.address_pool).
   if ! docker network inspect "${NETWORK_NAME}" &>/dev/null; then
     echo "Creating Docker network '${NETWORK_NAME}' (192.168.1.0/24)..."
     docker network create --subnet 192.168.1.0/24 --gateway 192.168.1.1 "${NETWORK_NAME}"
@@ -59,26 +61,6 @@ kubectl get nodes -o wide
 
 echo "::endgroup::"
 
-if [[ "${PROFILE}" == "platform" ]]; then
-  echo "::group::Create 'standard' StorageClass (workaround)"
-
-  # NIC's local provider hardcodes StorageClass "standard", but k3s only ships
-  # "local-path". Create a "standard" class backed by the same provisioner.
-  # TODO: Remove once nebari-dev/nebari-infrastructure-core#201 merges and
-  #       NIC supports configuring storage_class per provider.
-  kubectl apply -f - <<'SC'
-apiVersion: storage.k8s.io/v1
-kind: StorageClass
-metadata:
-  name: standard
-provisioner: rancher.io/local-path
-reclaimPolicy: Delete
-volumeBindingMode: WaitForFirstConsumer
-SC
-
-  echo "::endgroup::"
-fi
-
 # Set outputs
 KUBECONFIG_PATH="${HOME}/.kube/config"
 k3d kubeconfig merge "${CLUSTER_NAME}" --kubeconfig-merge-default
diff --git a/scripts/deploy-platform.sh b/scripts/deploy-platform.sh
index 6a747b1..76ff39e 100755
--- a/scripts/deploy-platform.sh
+++ b/scripts/deploy-platform.sh
@@ -21,6 +21,14 @@ cluster:
     # (not the top-level NebariConfig.kube_context, which is for "bring your
     # own cluster" mode that skips infra provisioning).
     kube_context: "k3d-${CLUSTER_NAME}"
+    # k3s ships "local-path" not "standard"; tell NIC to use it directly so we
+    # don't need to create a fake "standard" StorageClass as a workaround.
+    # Requires nebari-dev/nebari-infrastructure-core#201.
+    storage_class: local-path
+    # Explicitly declare the MetalLB pool so it matches the Docker network
+    # subnet (192.168.1.0/24) created in create-cluster.sh.
+    metallb:
+      address_pool: 192.168.1.100-192.168.1.110
     node_selectors:
       general:
         kubernetes.io/os: linux
diff --git a/scripts/install-nic.sh b/scripts/install-nic.sh
index 76f5685..3cd4e96 100755
--- a/scripts/install-nic.sh
+++ b/scripts/install-nic.sh
@@ -1,12 +1,17 @@
 #!/usr/bin/env bash
 set -euo pipefail
 
-# Build NIC from main. file:// git support (nebari-dev/nebari-infrastructure-core#136)
-# was merged into main on 2026-04-29 and the local_git branch was deleted.
+# Build NIC from the feature/local-configurable-infra-settings branch, which
+# makes StorageClass, MetalLB, and https_port configurable in the local provider
+# config (nebari-dev/nebari-infrastructure-core#201).
+#
+# This branch is pinned here to provide an integration test environment for
+# NIC PR #201 review. Once that PR merges into main (and eventually a new NIC
+# release ships), this should be updated back to main / a tagged release.
 # TODO: switch to downloading a pre-built binary once a NIC release ships
-#       that includes PR #136 (track: nebari-dev/action-nebari-sandbox#12).
+#       that includes PR #136 and PR #201 (track: nebari-dev/action-nebari-sandbox#12).
 NIC_REPO="https://github.com/nebari-dev/nebari-infrastructure-core.git"
-NIC_BRANCH="main"
+NIC_BRANCH="feature/local-configurable-infra-settings"
 NIC_SRC="/tmp/nebari-infrastructure-core"
 
 echo "::group::Install NIC (branch: ${NIC_BRANCH})"