diff --git a/.github/workflows/cd.yml b/.github/workflows/cd.yml
index ad14bf0..3053171 100644
--- a/.github/workflows/cd.yml
+++ b/.github/workflows/cd.yml
@@ -66,10 +66,18 @@ jobs:
           type=gha,mode=max
           type=registry,ref=${{ secrets.REGISTRY_REPOSITORY }}:buildcache,mode=max
         
+    - name: Extract SHA-suffixed tag
+      id: extract_tag
+      run: |
+        # Find the tag with SHA suffix (format: branch-sha)
+        TAGS='${{ steps.meta.outputs.tags }}'
+        SHA_TAG=$(echo "$TAGS" | grep -E '.*-[a-f0-9]{7}$' | head -1 | cut -d':' -f2)
+        echo $SHA_TAG >> $GITHUB_OUTPUT
+        
     - name: Deploy with Helm
       uses: ./.github/actions/helm-deploy
       with:
-        image_tag: ${{ fromJSON(steps.meta.outputs.json).tags[0] }}
+        image_tag: ${{ steps.extract_tag.outputs.tag }}
         registry_repository: ${{ secrets.REGISTRY_REPOSITORY }}
         kube_config_data: ${{ secrets.KUBE_CONFIG_DATA }}
         helm_values_env: ${{ secrets.HELM_VALUES_ENV }}
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index a0e5a68..1617854 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -42,6 +42,7 @@ jobs:
         flags: unittests
         fail_ci_if_error: false
         token: ${{ secrets.CODECOV_TOKEN }}
+        slug: XyLearningProgramming/slm_server
       
     - name: Lint with ruff
       run: |
diff --git a/deploy/helm/templates/deployment.yaml b/deploy/helm/templates/deployment.yaml
index f77662d..bd0b995 100644
--- a/deploy/helm/templates/deployment.yaml
+++ b/deploy/helm/templates/deployment.yaml
@@ -81,4 +81,6 @@ spec:
         - name: scripts
           configMap:
             name: {{ include "slm-server.fullname" . }}-scripts
-            defaultMode: 0755
\ No newline at end of file
+            defaultMode: 0755
+  strategy:
+    {{- toYaml .Values.strategy | nindent 4 }}
diff --git a/deploy/helm/values.yaml b/deploy/helm/values.yaml
index 890f6e8..2211e4c 100644
--- a/deploy/helm/values.yaml
+++ b/deploy/helm/values.yaml
@@ -105,3 +105,6 @@ probes:
     timeoutSeconds: 5
     successThreshold: 1
     failureThreshold: 3
+
+strategy:
+  type: Recreate