Merge branch 'main' into feat/intent-classification-lora-auto-detection

Author: yossiovadia

.github/workflows/ci-changes.yml

@@ -0,0 +1,88 @@
+name: CI Changes
+on:
+  workflow_call:
+    outputs:
+      core:
+        value: ${{ jobs.filter.outputs.core }}
+      dashboard:
+        value: ${{ jobs.filter.outputs.dashboard }}
+      website:
+        value: ${{ jobs.filter.outputs.website }}
+      docs:
+        value: ${{ jobs.filter.outputs.docs }}
+      helm:
+        value: ${{ jobs.filter.outputs.helm }}
+      e2e:
+        value: ${{ jobs.filter.outputs.e2e }}
+      ci:
+        value: ${{ jobs.filter.outputs.ci }}
+      docker:
+        value: ${{ jobs.filter.outputs.docker }}
+      make:
+        value: ${{ jobs.filter.outputs.make }}
+
+jobs:
+  filter:
+    runs-on: ubuntu-latest
+    outputs:
+      core: ${{ steps.changes.outputs.core }}
+      dashboard: ${{ steps.changes.outputs.dashboard }}
+      website: ${{ steps.changes.outputs.website }}
+      docs: ${{ steps.changes.outputs.docs }}
+      helm: ${{ steps.changes.outputs.helm }}
+      e2e: ${{ steps.changes.outputs.e2e }}
+      ci: ${{ steps.changes.outputs.ci }}
+      docker: ${{ steps.changes.outputs.docker }}
+      make: ${{ steps.changes.outputs.make }}
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+      - id: changes
+        uses: dorny/paths-filter@v3
+        with:
+          token: ${{ github.token }}
+          filters: |
+            core:
+              - 'src/**'
+              - 'candle-binding/**'
+              - 'bench/**'
+              - 'config/**'
+              - 'scripts/**'
+              - 'examples/**'
+              - 'e2e/**'
+              - 'tools/**'
+              - 'go.mod'
+              - 'go.sum'
+              - 'Cargo.toml'
+              - 'Cargo.lock'
+              - 'pyproject.toml'
+              - 'e2e-tests/llm-katan/**'
+            dashboard:
+              - 'dashboard/**'
+            website:
+              - 'website/**'
+            docs:
+              - 'docs/**'
+              - '**/*.md'
+              - '**/*.mdx'
+            helm:
+              - 'deploy/helm/**'
+            e2e:
+              - 'e2e-tests/**'
+              - 'e2e/**'
+              - 'scripts/quickstart.sh'
+              - 'deploy/docker-compose/**'
+              - 'deploy/kubernetes/**'
+              - 'deploy/kserve/**'
+              - 'deploy/openshift/**'
+            ci:
+              - '.github/workflows/**'
+            docker:
+              - 'Dockerfile*'
+              - 'docker/**'
+              - 'docker-compose*'
+              - 'deploy/docker-compose/**'
+            make:
+              - 'Makefile'
+              - 'tools/make/**'

.github/workflows/docker-publish.yml

@@ -29,17 +29,21 @@ on:
       - "src/**"
       - "e2e-tests/llm-katan/**"
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 jobs:
   # Build job for multi-architecture Docker images
   build_multiarch:
-    if: github.repository == 'vllm-project/semantic-router'
+    if: github.repository == 'vllm-project/semantic-router' && !github.event.pull_request.draft
     runs-on: ubuntu-latest
     permissions:
       contents: read
       packages: write
     strategy:
       matrix:
-        image: [extproc, llm-katan, dashboard]
+        image: [extproc, llm-katan, dashboard, model-downloader]
         # Multi-architecture build strategy:
         # - AMD64: Native build on ubuntu-latest (fast)
         # - ARM64: Cross-compilation on ubuntu-latest (faster than emulation)
@@ -171,6 +175,10 @@ jobs:
             echo "context=." >> $GITHUB_OUTPUT
             echo "dockerfile=./dashboard/backend/Dockerfile" >> $GITHUB_OUTPUT
             echo "platform=linux/${{ matrix.arch }}" >> $GITHUB_OUTPUT
+          elif [ "${{ matrix.image }}" = "model-downloader" ]; then
+            echo "context=." >> $GITHUB_OUTPUT
+            echo "dockerfile=./Dockerfile.model-downloader" >> $GITHUB_OUTPUT
+            echo "platform=linux/${{ matrix.arch }}" >> $GITHUB_OUTPUT
           fi
 
       - name: Generate tags
@@ -236,7 +244,7 @@ jobs:
       packages: write
     strategy:
       matrix:
-        image: [extproc, llm-katan, dashboard]
+        image: [extproc, llm-katan, dashboard, model-downloader]
 
     steps:
       - name: Log in to GitHub Container Registry

.github/workflows/docker-release.yml

@@ -5,6 +5,10 @@ on:
     tags:
       - 'v*'  # Triggers on version tags like v1.0.0, v2.1.3, etc.
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 jobs:
   build_and_push_extproc:
     if: github.repository == 'vllm-project/semantic-router'

.github/workflows/helm-publish.yml

@@ -2,15 +2,16 @@ name: Publish Helm Chart
 
 on:
   pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
     branches:
       - main
     paths:
-      - 'deploy/helm/**'
+      - "deploy/helm/**"
   push:
     branches:
       - main
     paths:
-      - 'deploy/helm/**'
+      - "deploy/helm/**"
   workflow_dispatch:
 
 env:
@@ -20,73 +21,61 @@ env:
   CHART_NAME: semantic-router
 
 jobs:
-  publish-chart:
-    name: Publish Helm Chart to GHCR
+  validate-and-package:
+    name: Validate and Package
+    if: ${{ !github.event.pull_request.draft }}
     runs-on: ubuntu-latest
     permissions:
       contents: read
-      packages: write
+    outputs:
+      chart-name: ${{ steps.package.outputs.name }}
     steps:
-      - name: Checkout code
-        uses: actions/checkout@v4
-
-      - name: Set up Helm
-        uses: azure/setup-helm@v4
+      - uses: actions/checkout@v4
+      - uses: azure/setup-helm@v4
         with:
           version: ${{ env.HELM_VERSION }}
 
-      - name: Set lowercase repository owner
-        run: echo "REPOSITORY_OWNER_LOWER=$(echo $GITHUB_REPOSITORY_OWNER | tr '[:upper:]' '[:lower:]')" >> $GITHUB_ENV
-
-      - name: Log in to GitHub Container Registry
-        run: |
-          echo "${{ secrets.GITHUB_TOKEN }}" | helm registry login ${{ env.REGISTRY }} --username ${{ github.actor }} --password-stdin
-
-      - name: Modify chart version to latest
-        run: |
-          echo "::group::Modify Chart.yaml"
-          # Backup original Chart.yaml
-          cp ${{ env.CHART_PATH }}/Chart.yaml ${{ env.CHART_PATH }}/Chart.yaml.bak
-
-          # Replace version with 'v0.0.0-latest' (valid semver with prerelease tag)
-          sed -i 's/^version:.*/version: v0.0.0-latest/' ${{ env.CHART_PATH }}/Chart.yaml
-
-          echo "Modified Chart.yaml:"
-          grep '^version:' ${{ env.CHART_PATH }}/Chart.yaml
-          echo "::endgroup::"
-
-      - name: Package Helm chart
-        run: |
-          echo "::group::Package Chart"
-          mkdir -p ./dist
-          helm package ${{ env.CHART_PATH }} --destination ./dist
-          echo "::endgroup::"
-
-          echo "::group::Packaged Chart"
-          ls -lh ./dist/
-          echo "::endgroup::"
+      - name: Lint chart
+        run: helm lint ${{ env.CHART_PATH }}
 
-      - name: Push chart to GHCR
+      - name: Package chart
+        id: package
         run: |
-          echo "::group::Push to GHCR"
-          CHART_PACKAGE="./dist/${{ env.CHART_NAME }}-v0.0.0-latest.tgz"
+          helm package ${{ env.CHART_PATH }} --version v0.0.0-latest --destination .
+          echo "name=$(ls *.tgz)" >> $GITHUB_OUTPUT
 
-          # Push to GHCR using OCI format with v0.0.0-latest tag
-          helm push "${CHART_PACKAGE}" oci://${{ env.REGISTRY }}/${{ env.REPOSITORY_OWNER_LOWER }}/charts
+      - name: Upload packaged chart
+        if: github.event_name != 'pull_request'
+        uses: actions/upload-artifact@v4
+        with:
+          name: helm-chart
+          path: "*.tgz"
+          retention-days: 1
 
-          echo "::endgroup::"
+  publish-chart:
+    name: Publish to GHCR
+    runs-on: ubuntu-latest
+    if: github.event_name != 'pull_request'
+    needs: validate-and-package
+    permissions:
+      packages: write
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: helm-chart
 
-          echo "✓ Chart published successfully!"
-          echo "Chart: ${{ env.REGISTRY }}/${{ env.REPOSITORY_OWNER_LOWER }}/charts/${{ env.CHART_NAME }}:v0.0.0-latest"
+      - uses: azure/setup-helm@v4
+        with:
+          version: ${{ env.HELM_VERSION }}
 
-      - name: Restore original Chart.yaml
-        if: always()
+      - name: Publish chart
+        env:
+          CHART: ${{ needs.validate-and-package.outputs.chart-name }}
+          REPO: ${{ env.REGISTRY }}/$(echo ${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]')/charts
         run: |
-          # Restore original Chart.yaml
-          if [ -f "${{ env.CHART_PATH }}/Chart.yaml.bak" ]; then
-            mv ${{ env.CHART_PATH }}/Chart.yaml.bak ${{ env.CHART_PATH }}/Chart.yaml
-            echo "Chart.yaml restored"
-          fi
+          echo "${{ secrets.GITHUB_TOKEN }}" | helm registry login ${{ env.REGISTRY }} -u ${{ github.actor }} --password-stdin
+          helm push "$CHART" "oci://$REPO"
+          helm registry logout ${{ env.REGISTRY }}
 
       - name: Create release summary
         run: |
@@ -118,9 +107,3 @@ jobs:
             --namespace vllm-semantic-router-system
           \`\`\`
           EOF
-
-      - name: Logout from registry
-        if: always()
-        run: |
-          helm registry logout ${{ env.REGISTRY }} || true
-

.github/workflows/integration-test-docker.yml

@@ -2,6 +2,7 @@ name: Integration Test [Docker Compose]
 
 on:
   pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
     branches:
       - main
     paths-ignore:
@@ -15,9 +16,13 @@ on:
       - '**/*.md'
   workflow_dispatch: # Allow manual triggering
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 jobs:
   test-quickstart:
-    if: github.repository == 'vllm-project/semantic-router'
+    if: github.repository == 'vllm-project/semantic-router' && !github.event.pull_request.draft
     runs-on: ubuntu-latest
     timeout-minutes: 30
 

.github/workflows/integration-test-dynamic-config.yml

@@ -2,6 +2,7 @@ name: Integration Test [Dynamic Config]
 
 on:
   pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
     branches:
       - main
     paths-ignore:
@@ -15,10 +16,15 @@ on:
       - '**/*.md'
   workflow_dispatch: # Allow manual triggering
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 jobs:
   integration-test:
+    if: ${{ !github.event.pull_request.draft }}
     runs-on: ubuntu-latest
-    timeout-minutes: 60
+    timeout-minutes: 180
 
     steps:
       - name: Check out the repo

.github/workflows/integration-test-helm.yml

@@ -8,13 +8,18 @@ on:
       - 'website/**'
       - '**/*.md'
   pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
     branches:
       - main
     paths-ignore:
       - 'website/**'
       - '**/*.md'
   workflow_dispatch:
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 env:
   HELM_VERSION: v3.14.0
   KIND_VERSION: v0.22.0
@@ -26,6 +31,7 @@ jobs:
   # Lint and validate Helm chart
   lint-chart:
     name: Lint Helm Chart
+    if: ${{ !github.event.pull_request.draft }}
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code