Added elasticsearch in conftest

.github/actions/install-llama-stack-client/action.yml

@@ -8,9 +8,6 @@ inputs:
     default: ""
 
 outputs:
-  uv-index-url:
-    description: 'UV_INDEX_URL to use (set for release branches)'
-    value: ${{ steps.configure.outputs.uv-index-url }}
   uv-extra-index-url:
     description: 'UV_EXTRA_INDEX_URL to use (set for release branches)'
     value: ${{ steps.configure.outputs.uv-extra-index-url }}
@@ -46,9 +43,8 @@ runs:
             exit 1
           fi
 
-          # Configure to use test.pypi for sync (to resolve RC versions)
-          echo "uv-index-url=https://test.pypi.org/simple/" >> $GITHUB_OUTPUT
-          echo "uv-extra-index-url=https://pypi.org/simple/" >> $GITHUB_OUTPUT
+          # Configure to use test.pypi as extra index (PyPI is primary)
+          echo "uv-extra-index-url=https://test.pypi.org/simple/" >> $GITHUB_OUTPUT
           echo "install-after-sync=true" >> $GITHUB_OUTPUT
           echo "install-source=git+https://github.com/llamastack/llama-stack-client-python.git@$BRANCH" >> $GITHUB_OUTPUT
         elif [ "${{ inputs.client-version }}" = "latest" ]; then

.github/actions/setup-runner/action.yml

@@ -27,9 +27,16 @@ runs:
     - name: Install dependencies
       shell: bash
       env:
-        UV_INDEX_URL: ${{ steps.client-config.outputs.uv-index-url }}
         UV_EXTRA_INDEX_URL: ${{ steps.client-config.outputs.uv-extra-index-url }}
       run: |
+        # Export UV env vars for current step and persist to GITHUB_ENV for subsequent steps
+        if [ -n "$UV_EXTRA_INDEX_URL" ]; then
+          export UV_INDEX_STRATEGY=unsafe-best-match
+          echo "UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL" >> $GITHUB_ENV
+          echo "UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY" >> $GITHUB_ENV
+          echo "Exported UV environment variables for current and subsequent steps"
+        fi
+
         echo "Updating project dependencies via uv sync"
         uv sync --all-groups
 

.github/mergify.yml

@@ -0,0 +1,23 @@
+pull_request_rules:
+- name: ping author on conflicts and add 'needs-rebase' label
+  conditions:
+      - conflict
+      - -closed
+  actions:
+    label:
+      add:
+        - needs-rebase
+    comment:
+      message: >
+       This pull request has merge conflicts that must be resolved before it
+       can be merged. @{{author}} please rebase it.
+       https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork
+
+- name: remove 'needs-rebase' label when conflict is resolved
+  conditions:
+      - -conflict
+      - -closed
+  actions:
+    label:
+      remove:
+        - needs-rebase

.github/workflows/README.md

@@ -13,7 +13,6 @@ Llama Stack uses GitHub Actions for Continuous Integration (CI). Below is a tabl
 | Integration Tests (Replay) | [integration-tests.yml](integration-tests.yml) | Run the integration test suites from tests/integration in replay mode |
 | Vector IO Integration Tests | [integration-vector-io-tests.yml](integration-vector-io-tests.yml) | Run the integration test suite with various VectorIO providers |
 | Pre-commit | [pre-commit.yml](pre-commit.yml) | Run pre-commit checks |
-| Pre-commit Bot | [precommit-trigger.yml](precommit-trigger.yml) | Pre-commit bot for PR |
 | Test Llama Stack Build | [providers-build.yml](providers-build.yml) | Test llama stack build |
 | Test llama stack list-deps | [providers-list-deps.yml](providers-list-deps.yml) | Test llama stack list-deps |
 | Python Package Build Test | [python-build-test.yml](python-build-test.yml) | Test building the llama-stack PyPI project |

.github/workflows/backward-compat.yml

@@ -37,7 +37,7 @@ jobs:
           python-version: '3.12'
 
       - name: Install uv
-        uses: astral-sh/setup-uv@6b9c6063abd6010835644d4c2e1bef4cf5cd0fca # v6.0.1
+        uses: astral-sh/setup-uv@85856786d1ce8acfbcc2f13a5f3fbd6b938f9f41 # v7.1.2
         with:
           enable-cache: true
 
@@ -415,7 +415,7 @@ jobs:
           python-version: '3.12'
 
       - name: Install uv
-        uses: astral-sh/setup-uv@6b9c6063abd6010835644d4c2e1bef4cf5cd0fca # v6.0.1
+        uses: astral-sh/setup-uv@85856786d1ce8acfbcc2f13a5f3fbd6b938f9f41 # v7.1.2
         with:
           enable-cache: true
 

.github/workflows/conformance.yml

@@ -22,7 +22,6 @@ on:
       - 'docs/static/stable-llama-stack-spec.yaml'       # Stable APIs spec
       - 'docs/static/experimental-llama-stack-spec.yaml' # Experimental APIs spec
       - 'docs/static/deprecated-llama-stack-spec.yaml'   # Deprecated APIs spec
-      - 'docs/static/llama-stack-spec.html'              # Legacy HTML spec
       - '.github/workflows/conformance.yml'              # This workflow itself
 
 concurrency:

.github/workflows/install-script-ci.yml

@@ -30,10 +30,16 @@ jobs:
 
       - name: Build a single provider
         run: |
+          BUILD_ARGS="--build-arg INSTALL_MODE=editable --build-arg DISTRO_NAME=starter"
+          if [ -n "${UV_EXTRA_INDEX_URL:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL"
+          fi
+          if [ -n "${UV_INDEX_STRATEGY:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY"
+          fi
           docker build . \
             -f containers/Containerfile \
-            --build-arg INSTALL_MODE=editable \
-            --build-arg DISTRO_NAME=starter \
+            $BUILD_ARGS \
             --tag llama-stack:starter-ci
 
       - name: Run installer end-to-end

.github/workflows/integration-tests.yml

@@ -22,6 +22,7 @@ on:
       - '.github/actions/setup-ollama/action.yml'
       - '.github/actions/setup-test-environment/action.yml'
       - '.github/actions/run-and-record-tests/action.yml'
+      - 'scripts/integration-tests.sh'
   schedule:
     # If changing the cron schedule, update the provider in the test-matrix job
     - cron: '0 0 * * *'  # (test latest client) Daily at 12 AM UTC

.github/workflows/pre-commit.yml

@@ -46,7 +46,7 @@ jobs:
           cache-dependency-path: 'src/llama_stack/ui/'
 
       - name: Set up uv
-        uses: astral-sh/setup-uv@2ddd2b9cb38ad8efd50337e8ab201519a34c9f24 # v7.1.1
+        uses: astral-sh/setup-uv@85856786d1ce8acfbcc2f13a5f3fbd6b938f9f41 # v7.1.2
 
       - name: Install npm dependencies
         run: npm ci
@@ -136,9 +136,12 @@ jobs:
 
       - name: Sync dev + type_checking dependencies
         env:
-          UV_INDEX_URL: ${{ steps.client-config.outputs.uv-index-url }}
           UV_EXTRA_INDEX_URL: ${{ steps.client-config.outputs.uv-extra-index-url }}
         run: |
+          if [ -n "$UV_EXTRA_INDEX_URL" ]; then
+            export UV_INDEX_STRATEGY="unsafe-best-match"
+          fi
+
           uv sync --group dev --group type_checking
 
           # Install specific client version after sync if needed
@@ -148,7 +151,13 @@ jobs:
           fi
 
       - name: Run mypy (full type_checking)
+        env:
+          UV_EXTRA_INDEX_URL: ${{ steps.client-config.outputs.uv-extra-index-url }}
         run: |
+          if [ -n "$UV_EXTRA_INDEX_URL" ]; then
+            export UV_INDEX_STRATEGY="unsafe-best-match"
+          fi
+
           set +e
           uv run --group dev --group type_checking mypy
           status=$?

.github/workflows/precommit-trigger.yml

@@ -1,227 +0,0 @@
-name: Pre-commit Bot
-
-run-name: Pre-commit bot for PR #${{ github.event.issue.number }}
-
-on:
-  issue_comment:
-    types: [created]
-
-jobs:
-  pre-commit:
-    # Only run on pull request comments
-    if: github.event.issue.pull_request && contains(github.event.comment.body, '@github-actions run precommit')
-    runs-on: ubuntu-latest
-    permissions:
-      contents: write
-      pull-requests: write
-
-    steps:
-      - name: Check comment author and get PR details
-        id: check_author
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            // Get PR details
-            const pr = await github.rest.pulls.get({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              pull_number: context.issue.number
-            });
-
-            // Check if commenter has write access or is the PR author
-            const commenter = context.payload.comment.user.login;
-            const prAuthor = pr.data.user.login;
-
-            let hasPermission = false;
-
-            // Check if commenter is PR author
-            if (commenter === prAuthor) {
-              hasPermission = true;
-              console.log(`Comment author ${commenter} is the PR author`);
-            } else {
-              // Check if commenter has write/admin access
-              try {
-                const permission = await github.rest.repos.getCollaboratorPermissionLevel({
-                  owner: context.repo.owner,
-                  repo: context.repo.repo,
-                  username: commenter
-                });
-
-                const level = permission.data.permission;
-                hasPermission = ['write', 'admin', 'maintain'].includes(level);
-                console.log(`Comment author ${commenter} has permission: ${level}`);
-              } catch (error) {
-                console.log(`Could not check permissions for ${commenter}: ${error.message}`);
-              }
-            }
-
-            if (!hasPermission) {
-              await github.rest.issues.createComment({
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                issue_number: context.issue.number,
-                body: `❌ @${commenter} You don't have permission to trigger pre-commit. Only PR authors or repository collaborators can run this command.`
-              });
-              core.setFailed(`User ${commenter} does not have permission`);
-              return;
-            }
-
-            // Save PR info for later steps
-            core.setOutput('pr_number', context.issue.number);
-            core.setOutput('pr_head_ref', pr.data.head.ref);
-            core.setOutput('pr_head_sha', pr.data.head.sha);
-            core.setOutput('pr_head_repo', pr.data.head.repo.full_name);
-            core.setOutput('pr_base_ref', pr.data.base.ref);
-            core.setOutput('is_fork', pr.data.head.repo.full_name !== context.payload.repository.full_name);
-            core.setOutput('authorized', 'true');
-
-      - name: React to comment
-        if: steps.check_author.outputs.authorized == 'true'
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            await github.rest.reactions.createForIssueComment({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              comment_id: context.payload.comment.id,
-              content: 'rocket'
-            });
-
-      - name: Comment starting
-        if: steps.check_author.outputs.authorized == 'true'
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            await github.rest.issues.createComment({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: ${{ steps.check_author.outputs.pr_number }},
-              body: `⏳ Running [pre-commit hooks](https://github.com/${context.repo.owner}/${context.repo.repo}/actions/runs/${context.runId}) on PR #${{ steps.check_author.outputs.pr_number }}...`
-            });
-
-      - name: Checkout PR branch (same-repo)
-        if: steps.check_author.outputs.authorized == 'true' && steps.check_author.outputs.is_fork == 'false'
-        uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
-        with:
-          ref: ${{ steps.check_author.outputs.pr_head_ref }}
-          fetch-depth: 0
-          token: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Checkout PR branch (fork)
-        if: steps.check_author.outputs.authorized == 'true' && steps.check_author.outputs.is_fork == 'true'
-        uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
-        with:
-          repository: ${{ steps.check_author.outputs.pr_head_repo }}
-          ref: ${{ steps.check_author.outputs.pr_head_ref }}
-          fetch-depth: 0
-          token: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Verify checkout
-        if: steps.check_author.outputs.authorized == 'true'
-        run: |
-          echo "Current SHA: $(git rev-parse HEAD)"
-          echo "Expected SHA: ${{ steps.check_author.outputs.pr_head_sha }}"
-          if [[ "$(git rev-parse HEAD)" != "${{ steps.check_author.outputs.pr_head_sha }}" ]]; then
-            echo "::error::Checked out SHA does not match expected SHA"
-            exit 1
-          fi
-
-      - name: Set up Python
-        if: steps.check_author.outputs.authorized == 'true'
-        uses: actions/setup-python@e797f83bcb11b83ae66e0230d6156d7c80228e7c # v6.0.0
-        with:
-          python-version: '3.12'
-          cache: pip
-          cache-dependency-path: |
-            **/requirements*.txt
-            .pre-commit-config.yaml
-
-      - name: Set up Node.js
-        if: steps.check_author.outputs.authorized == 'true'
-        uses: actions/setup-node@2028fbc5c25fe9cf00d9f06a71cc4710d4507903 # v6.0.0
-        with:
-          node-version: '20'
-          cache: 'npm'
-          cache-dependency-path: 'src/llama_stack/ui/'
-
-      - name: Install npm dependencies
-        if: steps.check_author.outputs.authorized == 'true'
-        run: npm ci
-        working-directory: src/llama_stack/ui
-
-      - name: Run pre-commit
-        if: steps.check_author.outputs.authorized == 'true'
-        id: precommit
-        uses: pre-commit/action@2c7b3805fd2a0fd8c1884dcaebf91fc102a13ecd # v3.0.1
-        continue-on-error: true
-        env:
-          SKIP: no-commit-to-branch
-          RUFF_OUTPUT_FORMAT: github
-
-      - name: Check for changes
-        if: steps.check_author.outputs.authorized == 'true'
-        id: changes
-        run: |
-          if ! git diff --exit-code || [ -n "$(git ls-files --others --exclude-standard)" ]; then
-            echo "has_changes=true" >> $GITHUB_OUTPUT
-            echo "Changes detected after pre-commit"
-          else
-            echo "has_changes=false" >> $GITHUB_OUTPUT
-            echo "No changes after pre-commit"
-          fi
-
-      - name: Commit and push changes
-        if: steps.check_author.outputs.authorized == 'true' && steps.changes.outputs.has_changes == 'true'
-        run: |
-          git config --local user.email "github-actions[bot]@users.noreply.github.com"
-          git config --local user.name "github-actions[bot]"
-
-          git add -A
-          git commit -m "style: apply pre-commit fixes
-
-          🤖 Applied by @github-actions bot via pre-commit workflow"
-
-          # Push changes
-          git push origin HEAD:${{ steps.check_author.outputs.pr_head_ref }}
-
-      - name: Comment success with changes
-        if: steps.check_author.outputs.authorized == 'true' && steps.changes.outputs.has_changes == 'true'
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            await github.rest.issues.createComment({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: ${{ steps.check_author.outputs.pr_number }},
-              body: `✅ Pre-commit hooks completed successfully!\n\n🔧 Changes have been committed and pushed to the PR branch.`
-            });
-
-      - name: Comment success without changes
-        if: steps.check_author.outputs.authorized == 'true' && steps.changes.outputs.has_changes == 'false' && steps.precommit.outcome == 'success'
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            await github.rest.issues.createComment({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: ${{ steps.check_author.outputs.pr_number }},
-              body: `✅ Pre-commit hooks passed!\n\n✨ No changes needed - your code is already formatted correctly.`
-            });
-
-      - name: Comment failure
-        if: failure()
-        uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
-        with:
-          github-token: ${{ secrets.GITHUB_TOKEN }}
-          script: |
-            await github.rest.issues.createComment({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: ${{ steps.check_author.outputs.pr_number }},
-              body: `❌ Pre-commit workflow failed!\n\nPlease check the [workflow logs](https://github.com/${context.repo.owner}/${context.repo.repo}/actions/runs/${context.runId}) for details.`
-            });

.github/workflows/providers-build.yml

@@ -72,10 +72,16 @@ jobs:
       - name: Build container image
         if: matrix.image-type == 'container'
         run: |
+          BUILD_ARGS="--build-arg INSTALL_MODE=editable --build-arg DISTRO_NAME=${{ matrix.distro }}"
+          if [ -n "${UV_EXTRA_INDEX_URL:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL"
+          fi
+          if [ -n "${UV_INDEX_STRATEGY:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY"
+          fi
           docker build . \
             -f containers/Containerfile \
-            --build-arg INSTALL_MODE=editable \
-            --build-arg DISTRO_NAME=${{ matrix.distro }} \
+            $BUILD_ARGS \
             --tag llama-stack:${{ matrix.distro }}-ci
 
       - name: Print dependencies in the image
@@ -108,12 +114,18 @@ jobs:
       - name: Build container image
         run: |
           BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "python:3.12-slim"' src/llama_stack/distributions/ci-tests/build.yaml)
+          BUILD_ARGS="--build-arg INSTALL_MODE=editable --build-arg DISTRO_NAME=ci-tests"
+          BUILD_ARGS="$BUILD_ARGS --build-arg BASE_IMAGE=$BASE_IMAGE"
+          BUILD_ARGS="$BUILD_ARGS --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml"
+          if [ -n "${UV_EXTRA_INDEX_URL:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL"
+          fi
+          if [ -n "${UV_INDEX_STRATEGY:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY"
+          fi
           docker build . \
             -f containers/Containerfile \
-            --build-arg INSTALL_MODE=editable \
-            --build-arg DISTRO_NAME=ci-tests \
-            --build-arg BASE_IMAGE="$BASE_IMAGE" \
-            --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml \
+            $BUILD_ARGS \
             -t llama-stack:ci-tests
 
       - name: Inspect the container image entrypoint
@@ -148,12 +160,18 @@ jobs:
       - name: Build UBI9 container image
         run: |
           BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "registry.access.redhat.com/ubi9:latest"' src/llama_stack/distributions/ci-tests/build.yaml)
+          BUILD_ARGS="--build-arg INSTALL_MODE=editable --build-arg DISTRO_NAME=ci-tests"
+          BUILD_ARGS="$BUILD_ARGS --build-arg BASE_IMAGE=$BASE_IMAGE"
+          BUILD_ARGS="$BUILD_ARGS --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml"
+          if [ -n "${UV_EXTRA_INDEX_URL:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL"
+          fi
+          if [ -n "${UV_INDEX_STRATEGY:-}" ]; then
+            BUILD_ARGS="$BUILD_ARGS --build-arg UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY"
+          fi
           docker build . \
             -f containers/Containerfile \
-            --build-arg INSTALL_MODE=editable \
-            --build-arg DISTRO_NAME=ci-tests \
-            --build-arg BASE_IMAGE="$BASE_IMAGE" \
-            --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml \
+            $BUILD_ARGS \
             -t llama-stack:ci-tests-ubi9
 
       - name: Inspect UBI9 image

.github/workflows/python-build-test.yml

@@ -24,7 +24,7 @@ jobs:
       uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
     - name: Install uv
-      uses: astral-sh/setup-uv@2ddd2b9cb38ad8efd50337e8ab201519a34c9f24 # v7.1.1
+      uses: astral-sh/setup-uv@85856786d1ce8acfbcc2f13a5f3fbd6b938f9f41 # v7.1.2
       with:
         python-version: ${{ matrix.python-version }}
         activate-environment: true

.pre-commit-config.yaml

@@ -52,10 +52,6 @@ repos:
         additional_dependencies:
         - black==24.3.0
 
--   repo: https://github.com/astral-sh/uv-pre-commit
-    rev: 0.7.20
-    hooks:
-    -   id: uv-lock
 
 -   repo: https://github.com/pre-commit/mirrors-mypy
     rev: v1.18.2
@@ -63,22 +59,13 @@ repos:
     -   id: mypy
         additional_dependencies:
           - uv==0.6.2
+          - mypy
           - pytest
           - rich
           - types-requests
           - pydantic
-          - httpx
         pass_filenames: false
 
--   repo: local
-    hooks:
-    -   id: mypy-full
-        name: mypy (full type_checking)
-        entry: uv run --group dev --group type_checking mypy
-        language: system
-        pass_filenames: false
-        stages: [manual]
-
 # - repo: https://github.com/tcort/markdown-link-check
 #   rev: v3.11.2
 #   hooks:
@@ -87,11 +74,26 @@ repos:
 
 -   repo: local
     hooks:
+      - id: uv-lock
+        name: uv-lock
+        additional_dependencies:
+          - uv==0.7.20
+        entry: ./scripts/uv-run-with-index.sh lock
+        language: python
+        pass_filenames: false
+        require_serial: true
+        files: ^(pyproject\.toml|uv\.lock)$
+      - id: mypy-full
+        name: mypy (full type_checking)
+        entry: ./scripts/uv-run-with-index.sh run --group dev --group type_checking mypy
+        language: system
+        pass_filenames: false
+        stages: [manual]
       - id: distro-codegen
         name: Distribution Template Codegen
         additional_dependencies:
           - uv==0.7.8
-        entry: uv run --group codegen ./scripts/distro_codegen.py
+        entry: ./scripts/uv-run-with-index.sh run --group codegen ./scripts/distro_codegen.py
         language: python
         pass_filenames: false
         require_serial: true
@@ -100,7 +102,7 @@ repos:
         name: Provider Codegen
         additional_dependencies:
           - uv==0.7.8
-        entry: uv run --group codegen ./scripts/provider_codegen.py
+        entry: ./scripts/uv-run-with-index.sh run --group codegen ./scripts/provider_codegen.py
         language: python
         pass_filenames: false
         require_serial: true
@@ -109,7 +111,7 @@ repos:
         name: API Spec Codegen
         additional_dependencies:
           - uv==0.7.8
-        entry: sh -c 'uv run ./docs/openapi_generator/run_openapi_generator.sh > /dev/null'
+        entry: sh -c './scripts/uv-run-with-index.sh run ./docs/openapi_generator/run_openapi_generator.sh > /dev/null'
         language: python
         pass_filenames: false
         require_serial: true
@@ -150,7 +152,7 @@ repos:
         name: Generate CI documentation
         additional_dependencies:
           - uv==0.7.8
-        entry: uv run ./scripts/gen-ci-docs.py
+        entry: ./scripts/uv-run-with-index.sh run ./scripts/gen-ci-docs.py
         language: python
         pass_filenames: false
         require_serial: true
@@ -162,6 +164,7 @@ repos:
         files: ^src/llama_stack/ui/.*\.(ts|tsx)$
         pass_filenames: false
         require_serial: true
+
       - id: check-log-usage
         name: Ensure 'llama_stack.log' usage for logging
         entry: bash
@@ -197,6 +200,7 @@ repos:
               echo;
               exit 1;
             } || true
+
 ci:
     autofix_commit_msg: 🎨 [pre-commit.ci] Auto format from pre-commit.com hooks
     autoupdate_commit_msg: ⬆ [pre-commit.ci] pre-commit autoupdate

client-sdks/stainless/openapi.yml

@@ -956,17 +956,32 @@ paths:
         List routes.
 
         List all available API routes with their methods and implementing providers.
-      parameters: []
+      parameters:
+        - name: api_filter
+          in: query
+          description: >-
+            Optional filter to control which routes are returned. Can be an API level
+            ('v1', 'v1alpha', 'v1beta') to show non-deprecated routes at that level,
+            or 'deprecated' to show deprecated routes across all levels. If not specified,
+            returns only non-deprecated v1 routes.
+          required: false
+          schema:
+            type: string
+            enum:
+              - v1
+              - v1alpha
+              - v1beta
+              - deprecated
       deprecated: false
   /v1/models:
     get:
       responses:
         '200':
-          description: A ListModelsResponse.
+          description: A OpenAIListModelsResponse.
           content:
             application/json:
               schema:
-                $ref: '#/components/schemas/ListModelsResponse'
+                $ref: '#/components/schemas/OpenAIListModelsResponse'
         '400':
           $ref: '#/components/responses/BadRequest400'
         '429':
@@ -979,8 +994,8 @@ paths:
           $ref: '#/components/responses/DefaultError'
       tags:
         - Models
-      summary: List all models.
-      description: List all models.
+      summary: List models using the OpenAI API.
+      description: List models using the OpenAI API.
       parameters: []
       deprecated: false
     post:
@@ -1970,40 +1985,6 @@ paths:
           schema:
             type: string
       deprecated: false
-  /v1/synthetic-data-generation/generate:
-    post:
-      responses:
-        '200':
-          description: >-
-            Response containing filtered synthetic data samples and optional statistics
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/SyntheticDataGenerationResponse'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - SyntheticDataGeneration (Coming Soon)
-      summary: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      description: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/SyntheticDataGenerateRequest'
-        required: true
-      deprecated: false
   /v1/tool-runtime/invoke:
     post:
       responses:
@@ -6817,6 +6798,88 @@ components:
       title: ListRoutesResponse
       description: >-
         Response containing a list of all available API routes.
+    OpenAIModel:
+      type: object
+      properties:
+        id:
+          type: string
+        object:
+          type: string
+          const: model
+          default: model
+        created:
+          type: integer
+        owned_by:
+          type: string
+        custom_metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+      additionalProperties: false
+      required:
+        - id
+        - object
+        - created
+        - owned_by
+      title: OpenAIModel
+      description: A model from OpenAI.
+    OpenAIListModelsResponse:
+      type: object
+      properties:
+        data:
+          type: array
+          items:
+            $ref: '#/components/schemas/OpenAIModel'
+      additionalProperties: false
+      required:
+        - data
+      title: OpenAIListModelsResponse
+    ModelType:
+      type: string
+      enum:
+        - llm
+        - embedding
+        - rerank
+      title: ModelType
+      description: >-
+        Enumeration of supported model types in Llama Stack.
+    RegisterModelRequest:
+      type: object
+      properties:
+        model_id:
+          type: string
+          description: The identifier of the model to register.
+        provider_model_id:
+          type: string
+          description: >-
+            The identifier of the model in the provider.
+        provider_id:
+          type: string
+          description: The identifier of the provider.
+        metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+          description: Any additional metadata for this model.
+        model_type:
+          $ref: '#/components/schemas/ModelType'
+          description: The type of model to register.
+      additionalProperties: false
+      required:
+        - model_id
+      title: RegisterModelRequest
     Model:
       type: object
       properties:
@@ -6874,57 +6937,6 @@ components:
       title: Model
       description: >-
         A model resource representing an AI model registered in Llama Stack.
-    ModelType:
-      type: string
-      enum:
-        - llm
-        - embedding
-        - rerank
-      title: ModelType
-      description: >-
-        Enumeration of supported model types in Llama Stack.
-    ListModelsResponse:
-      type: object
-      properties:
-        data:
-          type: array
-          items:
-            $ref: '#/components/schemas/Model'
-      additionalProperties: false
-      required:
-        - data
-      title: ListModelsResponse
-    RegisterModelRequest:
-      type: object
-      properties:
-        model_id:
-          type: string
-          description: The identifier of the model to register.
-        provider_model_id:
-          type: string
-          description: >-
-            The identifier of the model in the provider.
-        provider_id:
-          type: string
-          description: The identifier of the provider.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Any additional metadata for this model.
-        model_type:
-          $ref: '#/components/schemas/ModelType'
-          description: The type of model to register.
-      additionalProperties: false
-      required:
-        - model_id
-      title: RegisterModelRequest
     RunModerationRequest:
       type: object
       properties:
@@ -9938,45 +9950,29 @@ components:
       required:
         - shield_id
       title: RegisterShieldRequest
-    CompletionMessage:
+    InvokeToolRequest:
       type: object
       properties:
-        role:
+        tool_name:
           type: string
-          const: assistant
-          default: assistant
+          description: The name of the tool to invoke.
+        kwargs:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
           description: >-
-            Must be "assistant" to identify this as the model's response
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The content of the model's response
-        stop_reason:
-          type: string
-          enum:
-            - end_of_turn
-            - end_of_message
-            - out_of_tokens
-          description: >-
-            Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`:
-            The model finished generating the entire response. - `StopReason.end_of_message`:
-            The model finished generating but generated a partial response -- usually,
-            a tool call. The user may call the tool and continue the conversation
-            with the tool's response. - `StopReason.out_of_tokens`: The model ran
-            out of token budget.
-        tool_calls:
-          type: array
-          items:
-            $ref: '#/components/schemas/ToolCall'
-          description: >-
-            List of tool calls. Each tool call is a ToolCall object.
+            A dictionary of arguments to pass to the tool.
       additionalProperties: false
       required:
-        - role
-        - content
-        - stop_reason
-      title: CompletionMessage
-      description: >-
-        A message containing the model's (assistant) response in a chat conversation.
+        - tool_name
+        - kwargs
+      title: InvokeToolRequest
     ImageContentItem:
       type: object
       properties:
@@ -10023,41 +10019,6 @@ components:
         mapping:
           image: '#/components/schemas/ImageContentItem'
           text: '#/components/schemas/TextContentItem'
-    Message:
-      oneOf:
-        - $ref: '#/components/schemas/UserMessage'
-        - $ref: '#/components/schemas/SystemMessage'
-        - $ref: '#/components/schemas/ToolResponseMessage'
-        - $ref: '#/components/schemas/CompletionMessage'
-      discriminator:
-        propertyName: role
-        mapping:
-          user: '#/components/schemas/UserMessage'
-          system: '#/components/schemas/SystemMessage'
-          tool: '#/components/schemas/ToolResponseMessage'
-          assistant: '#/components/schemas/CompletionMessage'
-    SystemMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: system
-          default: system
-          description: >-
-            Must be "system" to identify this as a system message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the "system prompt". If multiple system messages are provided,
-            they are concatenated. The underlying Llama Stack code may also add other
-            system messages (for example, for formatting tool definitions).
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: SystemMessage
-      description: >-
-        A system message providing instructions or context to the model.
     TextContentItem:
       type: object
       properties:
@@ -10076,179 +10037,6 @@ components:
         - text
       title: TextContentItem
       description: A text content item
-    ToolCall:
-      type: object
-      properties:
-        call_id:
-          type: string
-        tool_name:
-          oneOf:
-            - type: string
-              enum:
-                - brave_search
-                - wolfram_alpha
-                - photogen
-                - code_interpreter
-              title: BuiltinTool
-            - type: string
-        arguments:
-          type: string
-      additionalProperties: false
-      required:
-        - call_id
-        - tool_name
-        - arguments
-      title: ToolCall
-    ToolResponseMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: tool
-          default: tool
-          description: >-
-            Must be "tool" to identify this as a tool response
-        call_id:
-          type: string
-          description: >-
-            Unique identifier for the tool call this response is for
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The response content from the tool
-      additionalProperties: false
-      required:
-        - role
-        - call_id
-        - content
-      title: ToolResponseMessage
-      description: >-
-        A message representing the result of a tool invocation.
-    URL:
-      type: object
-      properties:
-        uri:
-          type: string
-          description: The URL string pointing to the resource
-      additionalProperties: false
-      required:
-        - uri
-      title: URL
-      description: A URL reference to external content.
-    UserMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: user
-          default: user
-          description: >-
-            Must be "user" to identify this as a user message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the message, which can include text and other media
-        context:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) This field is used internally by Llama Stack to pass RAG context.
-            This field may be removed in the API in the future.
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: UserMessage
-      description: >-
-        A message from the user in a chat conversation.
-    SyntheticDataGenerateRequest:
-      type: object
-      properties:
-        dialogs:
-          type: array
-          items:
-            $ref: '#/components/schemas/Message'
-          description: >-
-            List of conversation messages to use as input for synthetic data generation
-        filtering_function:
-          type: string
-          enum:
-            - none
-            - random
-            - top_k
-            - top_p
-            - top_k_top_p
-            - sigmoid
-          description: >-
-            Type of filtering to apply to generated synthetic data samples
-        model:
-          type: string
-          description: >-
-            (Optional) The identifier of the model to use. The model must be registered
-            with Llama Stack and available via the /models endpoint
-      additionalProperties: false
-      required:
-        - dialogs
-        - filtering_function
-      title: SyntheticDataGenerateRequest
-    SyntheticDataGenerationResponse:
-      type: object
-      properties:
-        synthetic_data:
-          type: array
-          items:
-            type: object
-            additionalProperties:
-              oneOf:
-                - type: 'null'
-                - type: boolean
-                - type: number
-                - type: string
-                - type: array
-                - type: object
-          description: >-
-            List of generated synthetic data samples that passed the filtering criteria
-        statistics:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            (Optional) Statistical information about the generation process and filtering
-            results
-      additionalProperties: false
-      required:
-        - synthetic_data
-      title: SyntheticDataGenerationResponse
-      description: >-
-        Response from the synthetic data generation. Batch of (prompt, response, score)
-        tuples that pass the threshold.
-    InvokeToolRequest:
-      type: object
-      properties:
-        tool_name:
-          type: string
-          description: The name of the tool to invoke.
-        kwargs:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            A dictionary of arguments to pass to the tool.
-      additionalProperties: false
-      required:
-        - tool_name
-        - kwargs
-      title: InvokeToolRequest
     ToolInvocationResult:
       type: object
       properties:
@@ -10279,6 +10067,17 @@ components:
       additionalProperties: false
       title: ToolInvocationResult
       description: Result of a tool invocation.
+    URL:
+      type: object
+      properties:
+        uri:
+          type: string
+          description: The URL string pointing to the resource
+      additionalProperties: false
+      required:
+        - uri
+      title: URL
+      description: A URL reference to external content.
     ToolDef:
       type: object
       properties:
@@ -12311,6 +12110,45 @@ components:
       title: AgentSessionCreateResponse
       description: >-
         Response returned when creating a new agent session.
+    CompletionMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: assistant
+          default: assistant
+          description: >-
+            Must be "assistant" to identify this as the model's response
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: The content of the model's response
+        stop_reason:
+          type: string
+          enum:
+            - end_of_turn
+            - end_of_message
+            - out_of_tokens
+          description: >-
+            Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`:
+            The model finished generating the entire response. - `StopReason.end_of_message`:
+            The model finished generating but generated a partial response -- usually,
+            a tool call. The user may call the tool and continue the conversation
+            with the tool's response. - `StopReason.out_of_tokens`: The model ran
+            out of token budget.
+        tool_calls:
+          type: array
+          items:
+            $ref: '#/components/schemas/ToolCall'
+          description: >-
+            List of tool calls. Each tool call is a ToolCall object.
+      additionalProperties: false
+      required:
+        - role
+        - content
+        - stop_reason
+      title: CompletionMessage
+      description: >-
+        A message containing the model's (assistant) response in a chat conversation.
     InferenceStep:
       type: object
       properties:
@@ -12463,6 +12301,29 @@ components:
         - step_type
       title: ShieldCallStep
       description: A shield call step in an agent turn.
+    ToolCall:
+      type: object
+      properties:
+        call_id:
+          type: string
+        tool_name:
+          oneOf:
+            - type: string
+              enum:
+                - brave_search
+                - wolfram_alpha
+                - photogen
+                - code_interpreter
+              title: BuiltinTool
+            - type: string
+        arguments:
+          type: string
+      additionalProperties: false
+      required:
+        - call_id
+        - tool_name
+        - arguments
+      title: ToolCall
     ToolExecutionStep:
       type: object
       properties:
@@ -12550,6 +12411,30 @@ components:
         - content
       title: ToolResponse
       description: Response from a tool invocation.
+    ToolResponseMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: tool
+          default: tool
+          description: >-
+            Must be "tool" to identify this as a tool response
+        call_id:
+          type: string
+          description: >-
+            Unique identifier for the tool call this response is for
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: The response content from the tool
+      additionalProperties: false
+      required:
+        - role
+        - call_id
+        - content
+      title: ToolResponseMessage
+      description: >-
+        A message representing the result of a tool invocation.
     Turn:
       type: object
       properties:
@@ -12635,6 +12520,31 @@ components:
       title: Turn
       description: >-
         A single turn in an interaction with an Agentic System.
+    UserMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: user
+          default: user
+          description: >-
+            Must be "user" to identify this as a user message
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            The content of the message, which can include text and other media
+        context:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            (Optional) This field is used internally by Llama Stack to pass RAG context.
+            This field may be removed in the API in the future.
+      additionalProperties: false
+      required:
+        - role
+        - content
+      title: UserMessage
+      description: >-
+        A message from the user in a chat conversation.
     CreateAgentTurnRequest:
       type: object
       properties:
@@ -13248,6 +13158,28 @@ components:
         - sampling_params
       title: ModelCandidate
       description: A model candidate for evaluation.
+    SystemMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: system
+          default: system
+          description: >-
+            Must be "system" to identify this as a system message
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            The content of the "system prompt". If multiple system messages are provided,
+            they are concatenated. The underlying Llama Stack code may also add other
+            system messages (for example, for formatting tool definitions).
+      additionalProperties: false
+      required:
+        - role
+        - content
+      title: SystemMessage
+      description: >-
+        A system message providing instructions or context to the model.
     EvaluateRowsRequest:
       type: object
       properties:
@@ -14063,8 +13995,6 @@ tags:
     description: ''
   - name: Shields
     description: ''
-  - name: SyntheticDataGeneration (Coming Soon)
-    description: ''
   - name: ToolGroups
     description: ''
   - name: ToolRuntime
@@ -14092,7 +14022,6 @@ x-tagGroups:
       - Scoring
       - ScoringFunctions
       - Shields
-      - SyntheticDataGeneration (Coming Soon)
       - ToolGroups
       - ToolRuntime
       - VectorIO

containers/Containerfile

@@ -19,6 +19,8 @@ ARG KEEP_WORKSPACE=""
 ARG DISTRO_NAME="starter"
 ARG RUN_CONFIG_PATH=""
 ARG UV_HTTP_TIMEOUT=500
+ARG UV_EXTRA_INDEX_URL=""
+ARG UV_INDEX_STRATEGY=""
 ENV UV_HTTP_TIMEOUT=${UV_HTTP_TIMEOUT}
 ENV PYTHONDONTWRITEBYTECODE=1
 ENV PIP_DISABLE_PIP_VERSION_CHECK=1
@@ -45,7 +47,7 @@ RUN set -eux; \
         exit 1; \
     fi
 
-RUN pip install --no-cache uv
+RUN pip install --no-cache-dir uv
 ENV UV_SYSTEM_PYTHON=1
 
 ENV INSTALL_MODE=${INSTALL_MODE}
@@ -62,47 +64,60 @@ COPY . /workspace
 
 # Install the client package if it is provided
 # NOTE: this is installed before llama-stack since llama-stack depends on llama-stack-client-python
+# Unset UV index env vars to ensure we only use PyPI for the client
 RUN set -eux; \
+    unset UV_EXTRA_INDEX_URL UV_INDEX_STRATEGY; \
     if [ -n "$LLAMA_STACK_CLIENT_DIR" ]; then \
         if [ ! -d "$LLAMA_STACK_CLIENT_DIR" ]; then \
             echo "LLAMA_STACK_CLIENT_DIR is set but $LLAMA_STACK_CLIENT_DIR does not exist" >&2; \
             exit 1; \
         fi; \
-        uv pip install --no-cache -e "$LLAMA_STACK_CLIENT_DIR"; \
+        uv pip install --no-cache-dir -e "$LLAMA_STACK_CLIENT_DIR"; \
     fi;
 
 # Install llama-stack
+# Use UV_EXTRA_INDEX_URL inline only for editable install with RC dependencies
 RUN set -eux; \
+    SAVED_UV_EXTRA_INDEX_URL="${UV_EXTRA_INDEX_URL:-}"; \
+    SAVED_UV_INDEX_STRATEGY="${UV_INDEX_STRATEGY:-}"; \
+    unset UV_EXTRA_INDEX_URL UV_INDEX_STRATEGY; \
     if [ "$INSTALL_MODE" = "editable" ]; then \
         if [ ! -d "$LLAMA_STACK_DIR" ]; then \
             echo "INSTALL_MODE=editable requires LLAMA_STACK_DIR to point to a directory inside the build context" >&2; \
             exit 1; \
         fi; \
-        uv pip install --no-cache -e "$LLAMA_STACK_DIR"; \
-    elif [ "$INSTALL_MODE" = "test-pypi" ]; then \
-        uv pip install --no-cache fastapi libcst; \
-        if [ -n "$TEST_PYPI_VERSION" ]; then \
-            uv pip install --no-cache --extra-index-url https://test.pypi.org/simple/ --index-strategy unsafe-best-match "llama-stack==$TEST_PYPI_VERSION"; \
+        if [ -n "$SAVED_UV_EXTRA_INDEX_URL" ] && [ -n "$SAVED_UV_INDEX_STRATEGY" ]; then \
+            UV_EXTRA_INDEX_URL="$SAVED_UV_EXTRA_INDEX_URL" UV_INDEX_STRATEGY="$SAVED_UV_INDEX_STRATEGY" \
+                uv pip install --no-cache-dir -e "$LLAMA_STACK_DIR"; \
         else \
-            uv pip install --no-cache --extra-index-url https://test.pypi.org/simple/ --index-strategy unsafe-best-match llama-stack; \
+            uv pip install --no-cache-dir -e "$LLAMA_STACK_DIR"; \
+        fi; \
+    elif [ "$INSTALL_MODE" = "test-pypi" ]; then \
+        uv pip install --no-cache-dir fastapi libcst; \
+        if [ -n "$TEST_PYPI_VERSION" ]; then \
+            uv pip install --no-cache-dir --extra-index-url https://test.pypi.org/simple/ --index-strategy unsafe-best-match "llama-stack==$TEST_PYPI_VERSION"; \
+        else \
+            uv pip install --no-cache-dir --extra-index-url https://test.pypi.org/simple/ --index-strategy unsafe-best-match llama-stack; \
         fi; \
     else \
         if [ -n "$PYPI_VERSION" ]; then \
-            uv pip install --no-cache "llama-stack==$PYPI_VERSION"; \
+            uv pip install --no-cache-dir "llama-stack==$PYPI_VERSION"; \
         else \
-            uv pip install --no-cache llama-stack; \
+            uv pip install --no-cache-dir llama-stack; \
         fi; \
     fi;
 
 # Install the dependencies for the distribution
+# Explicitly unset UV index env vars to ensure we only use PyPI for distribution deps
 RUN set -eux; \
+    unset UV_EXTRA_INDEX_URL UV_INDEX_STRATEGY; \
     if [ -z "$DISTRO_NAME" ]; then \
         echo "DISTRO_NAME must be provided" >&2; \
         exit 1; \
     fi; \
     deps="$(llama stack list-deps "$DISTRO_NAME")"; \
     if [ -n "$deps" ]; then \
-        printf '%s\n' "$deps" | xargs -L1 uv pip install --no-cache; \
+        printf '%s\n' "$deps" | xargs -L1 uv pip install --no-cache-dir; \
     fi
 
 # Cleanup

docs/docs/concepts/apis/index.mdx

@@ -23,5 +23,4 @@ A Llama Stack API is described as a collection of REST endpoints. We currently s
 We are working on adding a few more APIs to complete the application lifecycle. These will include:
 - **Batch Inference**: run inference on a dataset of inputs
 - **Batch Agents**: run agents on a dataset of inputs
-- **Synthetic Data Generation**: generate synthetic data for model development
 - **Batches**: OpenAI-compatible batch management for inference

docs/docs/getting_started/detailed_tutorial.mdx

@@ -239,8 +239,13 @@ client = LlamaStackClient(base_url="http://localhost:8321")
 models = client.models.list()
 
 # Select the first LLM
-llm = next(m for m in models if m.model_type == "llm" and m.provider_id == "ollama")
-model_id = llm.identifier
+llm = next(
+    m for m in models
+    if m.custom_metadata
+    and m.custom_metadata.get("model_type") == "llm"
+    and m.custom_metadata.get("provider_id") == "ollama"
+)
+model_id = llm.id
 
 print("Model:", model_id)
 
@@ -279,8 +284,13 @@ import uuid
 client = LlamaStackClient(base_url=f"http://localhost:8321")
 
 models = client.models.list()
-llm = next(m for m in models if m.model_type == "llm" and m.provider_id == "ollama")
-model_id = llm.identifier
+llm = next(
+    m for m in models
+    if m.custom_metadata
+    and m.custom_metadata.get("model_type") == "llm"
+    and m.custom_metadata.get("provider_id") == "ollama"
+)
+model_id = llm.id
 
 agent = Agent(client, model=model_id, instructions="You are a helpful assistant.")
 
@@ -450,8 +460,11 @@ import uuid
 client = LlamaStackClient(base_url="http://localhost:8321")
 
 # Create a vector database instance
-embed_lm = next(m for m in client.models.list() if m.model_type == "embedding")
-embedding_model = embed_lm.identifier
+embed_lm = next(
+    m for m in client.models.list()
+    if m.custom_metadata and m.custom_metadata.get("model_type") == "embedding"
+)
+embedding_model = embed_lm.id
 vector_db_id = f"v{uuid.uuid4().hex}"
 # The VectorDB API is deprecated; the server now returns its own authoritative ID.
 # We capture the correct ID from the response's .identifier attribute.
@@ -489,9 +502,11 @@ client.tool_runtime.rag_tool.insert(
 llm = next(
     m
     for m in client.models.list()
-    if m.model_type == "llm" and m.provider_id == "ollama"
+    if m.custom_metadata
+    and m.custom_metadata.get("model_type") == "llm"
+    and m.custom_metadata.get("provider_id") == "ollama"
 )
-model = llm.identifier
+model = llm.id
 
 # Create the RAG agent
 rag_agent = Agent(

docs/openapi_generator/generate.py

@@ -84,7 +84,6 @@ def generate_spec(output_dir: Path, stability_filter: str = None, main_spec: boo
     )
 
     yaml_filename = f"{filename_prefix}llama-stack-spec.yaml"
-    html_filename = f"{filename_prefix}llama-stack-spec.html"
 
     with open(output_dir / yaml_filename, "w", encoding="utf-8") as fp:
         y = yaml.YAML()
@@ -102,11 +101,6 @@ def generate_spec(output_dir: Path, stability_filter: str = None, main_spec: boo
             fp,
         )
 
-    with open(output_dir / html_filename, "w") as fp:
-        spec.write_html(fp, pretty_print=True)
-
-    print(f"Generated {yaml_filename} and {html_filename}")
-
 def main(output_dir: str):
     output_dir = Path(output_dir)
     if not output_dir.exists():

docs/static/llama-stack-spec.html

@@ -1258,7 +1258,23 @@
                 ],
                 "summary": "List routes.",
                 "description": "List routes.\nList all available API routes with their methods and implementing providers.",
-                "parameters": [],
+                "parameters": [
+                    {
+                        "name": "api_filter",
+                        "in": "query",
+                        "description": "Optional filter to control which routes are returned. Can be an API level ('v1', 'v1alpha', 'v1beta') to show non-deprecated routes at that level, or 'deprecated' to show deprecated routes across all levels. If not specified, returns only non-deprecated v1 routes.",
+                        "required": false,
+                        "schema": {
+                            "type": "string",
+                            "enum": [
+                                "v1",
+                                "v1alpha",
+                                "v1beta",
+                                "deprecated"
+                            ]
+                        }
+                    }
+                ],
                 "deprecated": false
             }
         },
@@ -2634,51 +2650,6 @@
                 "deprecated": false
             }
         },
-        "/v1/synthetic-data-generation/generate": {
-            "post": {
-                "responses": {
-                    "200": {
-                        "description": "Response containing filtered synthetic data samples and optional statistics",
-                        "content": {
-                            "application/json": {
-                                "schema": {
-                                    "$ref": "#/components/schemas/SyntheticDataGenerationResponse"
-                                }
-                            }
-                        }
-                    },
-                    "400": {
-                        "$ref": "#/components/responses/BadRequest400"
-                    },
-                    "429": {
-                        "$ref": "#/components/responses/TooManyRequests429"
-                    },
-                    "500": {
-                        "$ref": "#/components/responses/InternalServerError500"
-                    },
-                    "default": {
-                        "$ref": "#/components/responses/DefaultError"
-                    }
-                },
-                "tags": [
-                    "SyntheticDataGeneration (Coming Soon)"
-                ],
-                "summary": "Generate synthetic data based on input dialogs and apply filtering.",
-                "description": "Generate synthetic data based on input dialogs and apply filtering.",
-                "parameters": [],
-                "requestBody": {
-                    "content": {
-                        "application/json": {
-                            "schema": {
-                                "$ref": "#/components/schemas/SyntheticDataGenerateRequest"
-                            }
-                        }
-                    },
-                    "required": true
-                },
-                "deprecated": false
-            }
-        },
         "/v1/tool-runtime/invoke": {
             "post": {
                 "responses": {
@@ -11462,44 +11433,46 @@
                 ],
                 "title": "RegisterShieldRequest"
             },
-            "CompletionMessage": {
+            "InvokeToolRequest": {
                 "type": "object",
                 "properties": {
-                    "role": {
+                    "tool_name": {
                         "type": "string",
-                        "const": "assistant",
-                        "default": "assistant",
-                        "description": "Must be \"assistant\" to identify this as the model's response"
+                        "description": "The name of the tool to invoke."
                     },
-                    "content": {
-                        "$ref": "#/components/schemas/InterleavedContent",
-                        "description": "The content of the model's response"
-                    },
-                    "stop_reason": {
-                        "type": "string",
-                        "enum": [
-                            "end_of_turn",
-                            "end_of_message",
-                            "out_of_tokens"
-                        ],
-                        "description": "Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`: The model finished generating the entire response. - `StopReason.end_of_message`: The model finished generating but generated a partial response -- usually, a tool call. The user may call the tool and continue the conversation with the tool's response. - `StopReason.out_of_tokens`: The model ran out of token budget."
-                    },
-                    "tool_calls": {
-                        "type": "array",
-                        "items": {
-                            "$ref": "#/components/schemas/ToolCall"
+                    "kwargs": {
+                        "type": "object",
+                        "additionalProperties": {
+                            "oneOf": [
+                                {
+                                    "type": "null"
+                                },
+                                {
+                                    "type": "boolean"
+                                },
+                                {
+                                    "type": "number"
+                                },
+                                {
+                                    "type": "string"
+                                },
+                                {
+                                    "type": "array"
+                                },
+                                {
+                                    "type": "object"
+                                }
+                            ]
                         },
-                        "description": "List of tool calls. Each tool call is a ToolCall object."
+                        "description": "A dictionary of arguments to pass to the tool."
                     }
                 },
                 "additionalProperties": false,
                 "required": [
-                    "role",
-                    "content",
-                    "stop_reason"
+                    "tool_name",
+                    "kwargs"
                 ],
-                "title": "CompletionMessage",
-                "description": "A message containing the model's (assistant) response in a chat conversation."
+                "title": "InvokeToolRequest"
             },
             "ImageContentItem": {
                 "type": "object",
@@ -11568,53 +11541,6 @@
                     }
                 }
             },
-            "Message": {
-                "oneOf": [
-                    {
-                        "$ref": "#/components/schemas/UserMessage"
-                    },
-                    {
-                        "$ref": "#/components/schemas/SystemMessage"
-                    },
-                    {
-                        "$ref": "#/components/schemas/ToolResponseMessage"
-                    },
-                    {
-                        "$ref": "#/components/schemas/CompletionMessage"
-                    }
-                ],
-                "discriminator": {
-                    "propertyName": "role",
-                    "mapping": {
-                        "user": "#/components/schemas/UserMessage",
-                        "system": "#/components/schemas/SystemMessage",
-                        "tool": "#/components/schemas/ToolResponseMessage",
-                        "assistant": "#/components/schemas/CompletionMessage"
-                    }
-                }
-            },
-            "SystemMessage": {
-                "type": "object",
-                "properties": {
-                    "role": {
-                        "type": "string",
-                        "const": "system",
-                        "default": "system",
-                        "description": "Must be \"system\" to identify this as a system message"
-                    },
-                    "content": {
-                        "$ref": "#/components/schemas/InterleavedContent",
-                        "description": "The content of the \"system prompt\". If multiple system messages are provided, they are concatenated. The underlying Llama Stack code may also add other system messages (for example, for formatting tool definitions)."
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "role",
-                    "content"
-                ],
-                "title": "SystemMessage",
-                "description": "A system message providing instructions or context to the model."
-            },
             "TextContentItem": {
                 "type": "object",
                 "properties": {
@@ -11637,250 +11563,6 @@
                 "title": "TextContentItem",
                 "description": "A text content item"
             },
-            "ToolCall": {
-                "type": "object",
-                "properties": {
-                    "call_id": {
-                        "type": "string"
-                    },
-                    "tool_name": {
-                        "oneOf": [
-                            {
-                                "type": "string",
-                                "enum": [
-                                    "brave_search",
-                                    "wolfram_alpha",
-                                    "photogen",
-                                    "code_interpreter"
-                                ],
-                                "title": "BuiltinTool"
-                            },
-                            {
-                                "type": "string"
-                            }
-                        ]
-                    },
-                    "arguments": {
-                        "type": "string"
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "call_id",
-                    "tool_name",
-                    "arguments"
-                ],
-                "title": "ToolCall"
-            },
-            "ToolResponseMessage": {
-                "type": "object",
-                "properties": {
-                    "role": {
-                        "type": "string",
-                        "const": "tool",
-                        "default": "tool",
-                        "description": "Must be \"tool\" to identify this as a tool response"
-                    },
-                    "call_id": {
-                        "type": "string",
-                        "description": "Unique identifier for the tool call this response is for"
-                    },
-                    "content": {
-                        "$ref": "#/components/schemas/InterleavedContent",
-                        "description": "The response content from the tool"
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "role",
-                    "call_id",
-                    "content"
-                ],
-                "title": "ToolResponseMessage",
-                "description": "A message representing the result of a tool invocation."
-            },
-            "URL": {
-                "type": "object",
-                "properties": {
-                    "uri": {
-                        "type": "string",
-                        "description": "The URL string pointing to the resource"
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "uri"
-                ],
-                "title": "URL",
-                "description": "A URL reference to external content."
-            },
-            "UserMessage": {
-                "type": "object",
-                "properties": {
-                    "role": {
-                        "type": "string",
-                        "const": "user",
-                        "default": "user",
-                        "description": "Must be \"user\" to identify this as a user message"
-                    },
-                    "content": {
-                        "$ref": "#/components/schemas/InterleavedContent",
-                        "description": "The content of the message, which can include text and other media"
-                    },
-                    "context": {
-                        "$ref": "#/components/schemas/InterleavedContent",
-                        "description": "(Optional) This field is used internally by Llama Stack to pass RAG context. This field may be removed in the API in the future."
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "role",
-                    "content"
-                ],
-                "title": "UserMessage",
-                "description": "A message from the user in a chat conversation."
-            },
-            "SyntheticDataGenerateRequest": {
-                "type": "object",
-                "properties": {
-                    "dialogs": {
-                        "type": "array",
-                        "items": {
-                            "$ref": "#/components/schemas/Message"
-                        },
-                        "description": "List of conversation messages to use as input for synthetic data generation"
-                    },
-                    "filtering_function": {
-                        "type": "string",
-                        "enum": [
-                            "none",
-                            "random",
-                            "top_k",
-                            "top_p",
-                            "top_k_top_p",
-                            "sigmoid"
-                        ],
-                        "description": "Type of filtering to apply to generated synthetic data samples"
-                    },
-                    "model": {
-                        "type": "string",
-                        "description": "(Optional) The identifier of the model to use. The model must be registered with Llama Stack and available via the /models endpoint"
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "dialogs",
-                    "filtering_function"
-                ],
-                "title": "SyntheticDataGenerateRequest"
-            },
-            "SyntheticDataGenerationResponse": {
-                "type": "object",
-                "properties": {
-                    "synthetic_data": {
-                        "type": "array",
-                        "items": {
-                            "type": "object",
-                            "additionalProperties": {
-                                "oneOf": [
-                                    {
-                                        "type": "null"
-                                    },
-                                    {
-                                        "type": "boolean"
-                                    },
-                                    {
-                                        "type": "number"
-                                    },
-                                    {
-                                        "type": "string"
-                                    },
-                                    {
-                                        "type": "array"
-                                    },
-                                    {
-                                        "type": "object"
-                                    }
-                                ]
-                            }
-                        },
-                        "description": "List of generated synthetic data samples that passed the filtering criteria"
-                    },
-                    "statistics": {
-                        "type": "object",
-                        "additionalProperties": {
-                            "oneOf": [
-                                {
-                                    "type": "null"
-                                },
-                                {
-                                    "type": "boolean"
-                                },
-                                {
-                                    "type": "number"
-                                },
-                                {
-                                    "type": "string"
-                                },
-                                {
-                                    "type": "array"
-                                },
-                                {
-                                    "type": "object"
-                                }
-                            ]
-                        },
-                        "description": "(Optional) Statistical information about the generation process and filtering results"
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "synthetic_data"
-                ],
-                "title": "SyntheticDataGenerationResponse",
-                "description": "Response from the synthetic data generation. Batch of (prompt, response, score) tuples that pass the threshold."
-            },
-            "InvokeToolRequest": {
-                "type": "object",
-                "properties": {
-                    "tool_name": {
-                        "type": "string",
-                        "description": "The name of the tool to invoke."
-                    },
-                    "kwargs": {
-                        "type": "object",
-                        "additionalProperties": {
-                            "oneOf": [
-                                {
-                                    "type": "null"
-                                },
-                                {
-                                    "type": "boolean"
-                                },
-                                {
-                                    "type": "number"
-                                },
-                                {
-                                    "type": "string"
-                                },
-                                {
-                                    "type": "array"
-                                },
-                                {
-                                    "type": "object"
-                                }
-                            ]
-                        },
-                        "description": "A dictionary of arguments to pass to the tool."
-                    }
-                },
-                "additionalProperties": false,
-                "required": [
-                    "tool_name",
-                    "kwargs"
-                ],
-                "title": "InvokeToolRequest"
-            },
             "ToolInvocationResult": {
                 "type": "object",
                 "properties": {
@@ -11927,6 +11609,21 @@
                 "title": "ToolInvocationResult",
                 "description": "Result of a tool invocation."
             },
+            "URL": {
+                "type": "object",
+                "properties": {
+                    "uri": {
+                        "type": "string",
+                        "description": "The URL string pointing to the resource"
+                    }
+                },
+                "additionalProperties": false,
+                "required": [
+                    "uri"
+                ],
+                "title": "URL",
+                "description": "A URL reference to external content."
+            },
             "ToolDef": {
                 "type": "object",
                 "properties": {
@@ -13977,10 +13674,6 @@
             "name": "Shields",
             "description": ""
         },
-        {
-            "name": "SyntheticDataGeneration (Coming Soon)",
-            "description": ""
-        },
         {
             "name": "ToolGroups",
             "description": ""
@@ -14011,7 +13704,6 @@
                 "Scoring",
                 "ScoringFunctions",
                 "Shields",
-                "SyntheticDataGeneration (Coming Soon)",
                 "ToolGroups",
                 "ToolRuntime",
                 "VectorIO"

docs/static/llama-stack-spec.yaml

@@ -953,17 +953,32 @@ paths:
         List routes.
 
         List all available API routes with their methods and implementing providers.
-      parameters: []
+      parameters:
+        - name: api_filter
+          in: query
+          description: >-
+            Optional filter to control which routes are returned. Can be an API level
+            ('v1', 'v1alpha', 'v1beta') to show non-deprecated routes at that level,
+            or 'deprecated' to show deprecated routes across all levels. If not specified,
+            returns only non-deprecated v1 routes.
+          required: false
+          schema:
+            type: string
+            enum:
+              - v1
+              - v1alpha
+              - v1beta
+              - deprecated
       deprecated: false
   /v1/models:
     get:
       responses:
         '200':
-          description: A ListModelsResponse.
+          description: A OpenAIListModelsResponse.
           content:
             application/json:
               schema:
-                $ref: '#/components/schemas/ListModelsResponse'
+                $ref: '#/components/schemas/OpenAIListModelsResponse'
         '400':
           $ref: '#/components/responses/BadRequest400'
         '429':
@@ -976,8 +991,8 @@ paths:
           $ref: '#/components/responses/DefaultError'
       tags:
         - Models
-      summary: List all models.
-      description: List all models.
+      summary: List models using the OpenAI API.
+      description: List models using the OpenAI API.
       parameters: []
       deprecated: false
     post:
@@ -1967,40 +1982,6 @@ paths:
           schema:
             type: string
       deprecated: false
-  /v1/synthetic-data-generation/generate:
-    post:
-      responses:
-        '200':
-          description: >-
-            Response containing filtered synthetic data samples and optional statistics
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/SyntheticDataGenerationResponse'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - SyntheticDataGeneration (Coming Soon)
-      summary: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      description: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/SyntheticDataGenerateRequest'
-        required: true
-      deprecated: false
   /v1/tool-runtime/invoke:
     post:
       responses:
@@ -5604,6 +5585,88 @@ components:
       title: ListRoutesResponse
       description: >-
         Response containing a list of all available API routes.
+    OpenAIModel:
+      type: object
+      properties:
+        id:
+          type: string
+        object:
+          type: string
+          const: model
+          default: model
+        created:
+          type: integer
+        owned_by:
+          type: string
+        custom_metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+      additionalProperties: false
+      required:
+        - id
+        - object
+        - created
+        - owned_by
+      title: OpenAIModel
+      description: A model from OpenAI.
+    OpenAIListModelsResponse:
+      type: object
+      properties:
+        data:
+          type: array
+          items:
+            $ref: '#/components/schemas/OpenAIModel'
+      additionalProperties: false
+      required:
+        - data
+      title: OpenAIListModelsResponse
+    ModelType:
+      type: string
+      enum:
+        - llm
+        - embedding
+        - rerank
+      title: ModelType
+      description: >-
+        Enumeration of supported model types in Llama Stack.
+    RegisterModelRequest:
+      type: object
+      properties:
+        model_id:
+          type: string
+          description: The identifier of the model to register.
+        provider_model_id:
+          type: string
+          description: >-
+            The identifier of the model in the provider.
+        provider_id:
+          type: string
+          description: The identifier of the provider.
+        metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+          description: Any additional metadata for this model.
+        model_type:
+          $ref: '#/components/schemas/ModelType'
+          description: The type of model to register.
+      additionalProperties: false
+      required:
+        - model_id
+      title: RegisterModelRequest
     Model:
       type: object
       properties:
@@ -5661,57 +5724,6 @@ components:
       title: Model
       description: >-
         A model resource representing an AI model registered in Llama Stack.
-    ModelType:
-      type: string
-      enum:
-        - llm
-        - embedding
-        - rerank
-      title: ModelType
-      description: >-
-        Enumeration of supported model types in Llama Stack.
-    ListModelsResponse:
-      type: object
-      properties:
-        data:
-          type: array
-          items:
-            $ref: '#/components/schemas/Model'
-      additionalProperties: false
-      required:
-        - data
-      title: ListModelsResponse
-    RegisterModelRequest:
-      type: object
-      properties:
-        model_id:
-          type: string
-          description: The identifier of the model to register.
-        provider_model_id:
-          type: string
-          description: >-
-            The identifier of the model in the provider.
-        provider_id:
-          type: string
-          description: The identifier of the provider.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Any additional metadata for this model.
-        model_type:
-          $ref: '#/components/schemas/ModelType'
-          description: The type of model to register.
-      additionalProperties: false
-      required:
-        - model_id
-      title: RegisterModelRequest
     RunModerationRequest:
       type: object
       properties:
@@ -8725,45 +8737,29 @@ components:
       required:
         - shield_id
       title: RegisterShieldRequest
-    CompletionMessage:
+    InvokeToolRequest:
       type: object
       properties:
-        role:
+        tool_name:
           type: string
-          const: assistant
-          default: assistant
+          description: The name of the tool to invoke.
+        kwargs:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
           description: >-
-            Must be "assistant" to identify this as the model's response
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The content of the model's response
-        stop_reason:
-          type: string
-          enum:
-            - end_of_turn
-            - end_of_message
-            - out_of_tokens
-          description: >-
-            Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`:
-            The model finished generating the entire response. - `StopReason.end_of_message`:
-            The model finished generating but generated a partial response -- usually,
-            a tool call. The user may call the tool and continue the conversation
-            with the tool's response. - `StopReason.out_of_tokens`: The model ran
-            out of token budget.
-        tool_calls:
-          type: array
-          items:
-            $ref: '#/components/schemas/ToolCall'
-          description: >-
-            List of tool calls. Each tool call is a ToolCall object.
+            A dictionary of arguments to pass to the tool.
       additionalProperties: false
       required:
-        - role
-        - content
-        - stop_reason
-      title: CompletionMessage
-      description: >-
-        A message containing the model's (assistant) response in a chat conversation.
+        - tool_name
+        - kwargs
+      title: InvokeToolRequest
     ImageContentItem:
       type: object
       properties:
@@ -8810,41 +8806,6 @@ components:
         mapping:
           image: '#/components/schemas/ImageContentItem'
           text: '#/components/schemas/TextContentItem'
-    Message:
-      oneOf:
-        - $ref: '#/components/schemas/UserMessage'
-        - $ref: '#/components/schemas/SystemMessage'
-        - $ref: '#/components/schemas/ToolResponseMessage'
-        - $ref: '#/components/schemas/CompletionMessage'
-      discriminator:
-        propertyName: role
-        mapping:
-          user: '#/components/schemas/UserMessage'
-          system: '#/components/schemas/SystemMessage'
-          tool: '#/components/schemas/ToolResponseMessage'
-          assistant: '#/components/schemas/CompletionMessage'
-    SystemMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: system
-          default: system
-          description: >-
-            Must be "system" to identify this as a system message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the "system prompt". If multiple system messages are provided,
-            they are concatenated. The underlying Llama Stack code may also add other
-            system messages (for example, for formatting tool definitions).
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: SystemMessage
-      description: >-
-        A system message providing instructions or context to the model.
     TextContentItem:
       type: object
       properties:
@@ -8863,179 +8824,6 @@ components:
         - text
       title: TextContentItem
       description: A text content item
-    ToolCall:
-      type: object
-      properties:
-        call_id:
-          type: string
-        tool_name:
-          oneOf:
-            - type: string
-              enum:
-                - brave_search
-                - wolfram_alpha
-                - photogen
-                - code_interpreter
-              title: BuiltinTool
-            - type: string
-        arguments:
-          type: string
-      additionalProperties: false
-      required:
-        - call_id
-        - tool_name
-        - arguments
-      title: ToolCall
-    ToolResponseMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: tool
-          default: tool
-          description: >-
-            Must be "tool" to identify this as a tool response
-        call_id:
-          type: string
-          description: >-
-            Unique identifier for the tool call this response is for
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The response content from the tool
-      additionalProperties: false
-      required:
-        - role
-        - call_id
-        - content
-      title: ToolResponseMessage
-      description: >-
-        A message representing the result of a tool invocation.
-    URL:
-      type: object
-      properties:
-        uri:
-          type: string
-          description: The URL string pointing to the resource
-      additionalProperties: false
-      required:
-        - uri
-      title: URL
-      description: A URL reference to external content.
-    UserMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: user
-          default: user
-          description: >-
-            Must be "user" to identify this as a user message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the message, which can include text and other media
-        context:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) This field is used internally by Llama Stack to pass RAG context.
-            This field may be removed in the API in the future.
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: UserMessage
-      description: >-
-        A message from the user in a chat conversation.
-    SyntheticDataGenerateRequest:
-      type: object
-      properties:
-        dialogs:
-          type: array
-          items:
-            $ref: '#/components/schemas/Message'
-          description: >-
-            List of conversation messages to use as input for synthetic data generation
-        filtering_function:
-          type: string
-          enum:
-            - none
-            - random
-            - top_k
-            - top_p
-            - top_k_top_p
-            - sigmoid
-          description: >-
-            Type of filtering to apply to generated synthetic data samples
-        model:
-          type: string
-          description: >-
-            (Optional) The identifier of the model to use. The model must be registered
-            with Llama Stack and available via the /models endpoint
-      additionalProperties: false
-      required:
-        - dialogs
-        - filtering_function
-      title: SyntheticDataGenerateRequest
-    SyntheticDataGenerationResponse:
-      type: object
-      properties:
-        synthetic_data:
-          type: array
-          items:
-            type: object
-            additionalProperties:
-              oneOf:
-                - type: 'null'
-                - type: boolean
-                - type: number
-                - type: string
-                - type: array
-                - type: object
-          description: >-
-            List of generated synthetic data samples that passed the filtering criteria
-        statistics:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            (Optional) Statistical information about the generation process and filtering
-            results
-      additionalProperties: false
-      required:
-        - synthetic_data
-      title: SyntheticDataGenerationResponse
-      description: >-
-        Response from the synthetic data generation. Batch of (prompt, response, score)
-        tuples that pass the threshold.
-    InvokeToolRequest:
-      type: object
-      properties:
-        tool_name:
-          type: string
-          description: The name of the tool to invoke.
-        kwargs:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            A dictionary of arguments to pass to the tool.
-      additionalProperties: false
-      required:
-        - tool_name
-        - kwargs
-      title: InvokeToolRequest
     ToolInvocationResult:
       type: object
       properties:
@@ -9066,6 +8854,17 @@ components:
       additionalProperties: false
       title: ToolInvocationResult
       description: Result of a tool invocation.
+    URL:
+      type: object
+      properties:
+        uri:
+          type: string
+          description: The URL string pointing to the resource
+      additionalProperties: false
+      required:
+        - uri
+      title: URL
+      description: A URL reference to external content.
     ToolDef:
       type: object
       properties:
@@ -10667,8 +10466,6 @@ tags:
     description: ''
   - name: Shields
     description: ''
-  - name: SyntheticDataGeneration (Coming Soon)
-    description: ''
   - name: ToolGroups
     description: ''
   - name: ToolRuntime
@@ -10691,7 +10488,6 @@ x-tagGroups:
       - Scoring
       - ScoringFunctions
       - Shields
-      - SyntheticDataGeneration (Coming Soon)
       - ToolGroups
       - ToolRuntime
       - VectorIO

docs/static/stainless-llama-stack-spec.yaml

@@ -956,17 +956,32 @@ paths:
         List routes.
 
         List all available API routes with their methods and implementing providers.
-      parameters: []
+      parameters:
+        - name: api_filter
+          in: query
+          description: >-
+            Optional filter to control which routes are returned. Can be an API level
+            ('v1', 'v1alpha', 'v1beta') to show non-deprecated routes at that level,
+            or 'deprecated' to show deprecated routes across all levels. If not specified,
+            returns only non-deprecated v1 routes.
+          required: false
+          schema:
+            type: string
+            enum:
+              - v1
+              - v1alpha
+              - v1beta
+              - deprecated
       deprecated: false
   /v1/models:
     get:
       responses:
         '200':
-          description: A ListModelsResponse.
+          description: A OpenAIListModelsResponse.
           content:
             application/json:
               schema:
-                $ref: '#/components/schemas/ListModelsResponse'
+                $ref: '#/components/schemas/OpenAIListModelsResponse'
         '400':
           $ref: '#/components/responses/BadRequest400'
         '429':
@@ -979,8 +994,8 @@ paths:
           $ref: '#/components/responses/DefaultError'
       tags:
         - Models
-      summary: List all models.
-      description: List all models.
+      summary: List models using the OpenAI API.
+      description: List models using the OpenAI API.
       parameters: []
       deprecated: false
     post:
@@ -1970,40 +1985,6 @@ paths:
           schema:
             type: string
       deprecated: false
-  /v1/synthetic-data-generation/generate:
-    post:
-      responses:
-        '200':
-          description: >-
-            Response containing filtered synthetic data samples and optional statistics
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/SyntheticDataGenerationResponse'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - SyntheticDataGeneration (Coming Soon)
-      summary: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      description: >-
-        Generate synthetic data based on input dialogs and apply filtering.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/SyntheticDataGenerateRequest'
-        required: true
-      deprecated: false
   /v1/tool-runtime/invoke:
     post:
       responses:
@@ -6817,6 +6798,88 @@ components:
       title: ListRoutesResponse
       description: >-
         Response containing a list of all available API routes.
+    OpenAIModel:
+      type: object
+      properties:
+        id:
+          type: string
+        object:
+          type: string
+          const: model
+          default: model
+        created:
+          type: integer
+        owned_by:
+          type: string
+        custom_metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+      additionalProperties: false
+      required:
+        - id
+        - object
+        - created
+        - owned_by
+      title: OpenAIModel
+      description: A model from OpenAI.
+    OpenAIListModelsResponse:
+      type: object
+      properties:
+        data:
+          type: array
+          items:
+            $ref: '#/components/schemas/OpenAIModel'
+      additionalProperties: false
+      required:
+        - data
+      title: OpenAIListModelsResponse
+    ModelType:
+      type: string
+      enum:
+        - llm
+        - embedding
+        - rerank
+      title: ModelType
+      description: >-
+        Enumeration of supported model types in Llama Stack.
+    RegisterModelRequest:
+      type: object
+      properties:
+        model_id:
+          type: string
+          description: The identifier of the model to register.
+        provider_model_id:
+          type: string
+          description: >-
+            The identifier of the model in the provider.
+        provider_id:
+          type: string
+          description: The identifier of the provider.
+        metadata:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
+          description: Any additional metadata for this model.
+        model_type:
+          $ref: '#/components/schemas/ModelType'
+          description: The type of model to register.
+      additionalProperties: false
+      required:
+        - model_id
+      title: RegisterModelRequest
     Model:
       type: object
       properties:
@@ -6874,57 +6937,6 @@ components:
       title: Model
       description: >-
         A model resource representing an AI model registered in Llama Stack.
-    ModelType:
-      type: string
-      enum:
-        - llm
-        - embedding
-        - rerank
-      title: ModelType
-      description: >-
-        Enumeration of supported model types in Llama Stack.
-    ListModelsResponse:
-      type: object
-      properties:
-        data:
-          type: array
-          items:
-            $ref: '#/components/schemas/Model'
-      additionalProperties: false
-      required:
-        - data
-      title: ListModelsResponse
-    RegisterModelRequest:
-      type: object
-      properties:
-        model_id:
-          type: string
-          description: The identifier of the model to register.
-        provider_model_id:
-          type: string
-          description: >-
-            The identifier of the model in the provider.
-        provider_id:
-          type: string
-          description: The identifier of the provider.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Any additional metadata for this model.
-        model_type:
-          $ref: '#/components/schemas/ModelType'
-          description: The type of model to register.
-      additionalProperties: false
-      required:
-        - model_id
-      title: RegisterModelRequest
     RunModerationRequest:
       type: object
       properties:
@@ -9938,45 +9950,29 @@ components:
       required:
         - shield_id
       title: RegisterShieldRequest
-    CompletionMessage:
+    InvokeToolRequest:
       type: object
       properties:
-        role:
+        tool_name:
           type: string
-          const: assistant
-          default: assistant
+          description: The name of the tool to invoke.
+        kwargs:
+          type: object
+          additionalProperties:
+            oneOf:
+              - type: 'null'
+              - type: boolean
+              - type: number
+              - type: string
+              - type: array
+              - type: object
           description: >-
-            Must be "assistant" to identify this as the model's response
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The content of the model's response
-        stop_reason:
-          type: string
-          enum:
-            - end_of_turn
-            - end_of_message
-            - out_of_tokens
-          description: >-
-            Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`:
-            The model finished generating the entire response. - `StopReason.end_of_message`:
-            The model finished generating but generated a partial response -- usually,
-            a tool call. The user may call the tool and continue the conversation
-            with the tool's response. - `StopReason.out_of_tokens`: The model ran
-            out of token budget.
-        tool_calls:
-          type: array
-          items:
-            $ref: '#/components/schemas/ToolCall'
-          description: >-
-            List of tool calls. Each tool call is a ToolCall object.
+            A dictionary of arguments to pass to the tool.
       additionalProperties: false
       required:
-        - role
-        - content
-        - stop_reason
-      title: CompletionMessage
-      description: >-
-        A message containing the model's (assistant) response in a chat conversation.
+        - tool_name
+        - kwargs
+      title: InvokeToolRequest
     ImageContentItem:
       type: object
       properties:
@@ -10023,41 +10019,6 @@ components:
         mapping:
           image: '#/components/schemas/ImageContentItem'
           text: '#/components/schemas/TextContentItem'
-    Message:
-      oneOf:
-        - $ref: '#/components/schemas/UserMessage'
-        - $ref: '#/components/schemas/SystemMessage'
-        - $ref: '#/components/schemas/ToolResponseMessage'
-        - $ref: '#/components/schemas/CompletionMessage'
-      discriminator:
-        propertyName: role
-        mapping:
-          user: '#/components/schemas/UserMessage'
-          system: '#/components/schemas/SystemMessage'
-          tool: '#/components/schemas/ToolResponseMessage'
-          assistant: '#/components/schemas/CompletionMessage'
-    SystemMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: system
-          default: system
-          description: >-
-            Must be "system" to identify this as a system message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the "system prompt". If multiple system messages are provided,
-            they are concatenated. The underlying Llama Stack code may also add other
-            system messages (for example, for formatting tool definitions).
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: SystemMessage
-      description: >-
-        A system message providing instructions or context to the model.
     TextContentItem:
       type: object
       properties:
@@ -10076,179 +10037,6 @@ components:
         - text
       title: TextContentItem
       description: A text content item
-    ToolCall:
-      type: object
-      properties:
-        call_id:
-          type: string
-        tool_name:
-          oneOf:
-            - type: string
-              enum:
-                - brave_search
-                - wolfram_alpha
-                - photogen
-                - code_interpreter
-              title: BuiltinTool
-            - type: string
-        arguments:
-          type: string
-      additionalProperties: false
-      required:
-        - call_id
-        - tool_name
-        - arguments
-      title: ToolCall
-    ToolResponseMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: tool
-          default: tool
-          description: >-
-            Must be "tool" to identify this as a tool response
-        call_id:
-          type: string
-          description: >-
-            Unique identifier for the tool call this response is for
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: The response content from the tool
-      additionalProperties: false
-      required:
-        - role
-        - call_id
-        - content
-      title: ToolResponseMessage
-      description: >-
-        A message representing the result of a tool invocation.
-    URL:
-      type: object
-      properties:
-        uri:
-          type: string
-          description: The URL string pointing to the resource
-      additionalProperties: false
-      required:
-        - uri
-      title: URL
-      description: A URL reference to external content.
-    UserMessage:
-      type: object
-      properties:
-        role:
-          type: string
-          const: user
-          default: user
-          description: >-
-            Must be "user" to identify this as a user message
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The content of the message, which can include text and other media
-        context:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) This field is used internally by Llama Stack to pass RAG context.
-            This field may be removed in the API in the future.
-      additionalProperties: false
-      required:
-        - role
-        - content
-      title: UserMessage
-      description: >-
-        A message from the user in a chat conversation.
-    SyntheticDataGenerateRequest:
-      type: object
-      properties:
-        dialogs:
-          type: array
-          items:
-            $ref: '#/components/schemas/Message'
-          description: >-
-            List of conversation messages to use as input for synthetic data generation
-        filtering_function:
-          type: string
-          enum:
-            - none
-            - random
-            - top_k
-            - top_p
-            - top_k_top_p
-            - sigmoid
-          description: >-
-            Type of filtering to apply to generated synthetic data samples
-        model:
-          type: string
-          description: >-
-            (Optional) The identifier of the model to use. The model must be registered
-            with Llama Stack and available via the /models endpoint
-      additionalProperties: false
-      required:
-        - dialogs
-        - filtering_function
-      title: SyntheticDataGenerateRequest
-    SyntheticDataGenerationResponse:
-      type: object
-      properties:
-        synthetic_data:
-          type: array
-          items:
-            type: object
-            additionalProperties:
-              oneOf:
-                - type: 'null'
-                - type: boolean
-                - type: number
-                - type: string
-                - type: array
-                - type: object
-          description: >-
-            List of generated synthetic data samples that passed the filtering criteria
-        statistics:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            (Optional) Statistical information about the generation process and filtering
-            results
-      additionalProperties: false
-      required:
-        - synthetic_data
-      title: SyntheticDataGenerationResponse
-      description: >-
-        Response from the synthetic data generation. Batch of (prompt, response, score)
-        tuples that pass the threshold.
-    InvokeToolRequest:
-      type: object
-      properties:
-        tool_name:
-          type: string
-          description: The name of the tool to invoke.
-        kwargs:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            A dictionary of arguments to pass to the tool.
-      additionalProperties: false
-      required:
-        - tool_name
-        - kwargs
-      title: InvokeToolRequest
     ToolInvocationResult:
       type: object
       properties:
@@ -10279,6 +10067,17 @@ components:
       additionalProperties: false
       title: ToolInvocationResult
       description: Result of a tool invocation.
+    URL:
+      type: object
+      properties:
+        uri:
+          type: string
+          description: The URL string pointing to the resource
+      additionalProperties: false
+      required:
+        - uri
+      title: URL
+      description: A URL reference to external content.
     ToolDef:
       type: object
       properties:
@@ -12311,6 +12110,45 @@ components:
       title: AgentSessionCreateResponse
       description: >-
         Response returned when creating a new agent session.
+    CompletionMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: assistant
+          default: assistant
+          description: >-
+            Must be "assistant" to identify this as the model's response
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: The content of the model's response
+        stop_reason:
+          type: string
+          enum:
+            - end_of_turn
+            - end_of_message
+            - out_of_tokens
+          description: >-
+            Reason why the model stopped generating. Options are: - `StopReason.end_of_turn`:
+            The model finished generating the entire response. - `StopReason.end_of_message`:
+            The model finished generating but generated a partial response -- usually,
+            a tool call. The user may call the tool and continue the conversation
+            with the tool's response. - `StopReason.out_of_tokens`: The model ran
+            out of token budget.
+        tool_calls:
+          type: array
+          items:
+            $ref: '#/components/schemas/ToolCall'
+          description: >-
+            List of tool calls. Each tool call is a ToolCall object.
+      additionalProperties: false
+      required:
+        - role
+        - content
+        - stop_reason
+      title: CompletionMessage
+      description: >-
+        A message containing the model's (assistant) response in a chat conversation.
     InferenceStep:
       type: object
       properties:
@@ -12463,6 +12301,29 @@ components:
         - step_type
       title: ShieldCallStep
       description: A shield call step in an agent turn.
+    ToolCall:
+      type: object
+      properties:
+        call_id:
+          type: string
+        tool_name:
+          oneOf:
+            - type: string
+              enum:
+                - brave_search
+                - wolfram_alpha
+                - photogen
+                - code_interpreter
+              title: BuiltinTool
+            - type: string
+        arguments:
+          type: string
+      additionalProperties: false
+      required:
+        - call_id
+        - tool_name
+        - arguments
+      title: ToolCall
     ToolExecutionStep:
       type: object
       properties:
@@ -12550,6 +12411,30 @@ components:
         - content
       title: ToolResponse
       description: Response from a tool invocation.
+    ToolResponseMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: tool
+          default: tool
+          description: >-
+            Must be "tool" to identify this as a tool response
+        call_id:
+          type: string
+          description: >-
+            Unique identifier for the tool call this response is for
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: The response content from the tool
+      additionalProperties: false
+      required:
+        - role
+        - call_id
+        - content
+      title: ToolResponseMessage
+      description: >-
+        A message representing the result of a tool invocation.
     Turn:
       type: object
       properties:
@@ -12635,6 +12520,31 @@ components:
       title: Turn
       description: >-
         A single turn in an interaction with an Agentic System.
+    UserMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: user
+          default: user
+          description: >-
+            Must be "user" to identify this as a user message
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            The content of the message, which can include text and other media
+        context:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            (Optional) This field is used internally by Llama Stack to pass RAG context.
+            This field may be removed in the API in the future.
+      additionalProperties: false
+      required:
+        - role
+        - content
+      title: UserMessage
+      description: >-
+        A message from the user in a chat conversation.
     CreateAgentTurnRequest:
       type: object
       properties:
@@ -13248,6 +13158,28 @@ components:
         - sampling_params
       title: ModelCandidate
       description: A model candidate for evaluation.
+    SystemMessage:
+      type: object
+      properties:
+        role:
+          type: string
+          const: system
+          default: system
+          description: >-
+            Must be "system" to identify this as a system message
+        content:
+          $ref: '#/components/schemas/InterleavedContent'
+          description: >-
+            The content of the "system prompt". If multiple system messages are provided,
+            they are concatenated. The underlying Llama Stack code may also add other
+            system messages (for example, for formatting tool definitions).
+      additionalProperties: false
+      required:
+        - role
+        - content
+      title: SystemMessage
+      description: >-
+        A system message providing instructions or context to the model.
     EvaluateRowsRequest:
       type: object
       properties:
@@ -14063,8 +13995,6 @@ tags:
     description: ''
   - name: Shields
     description: ''
-  - name: SyntheticDataGeneration (Coming Soon)
-    description: ''
   - name: ToolGroups
     description: ''
   - name: ToolRuntime
@@ -14092,7 +14022,6 @@ x-tagGroups:
       - Scoring
       - ScoringFunctions
       - Shields
-      - SyntheticDataGeneration (Coming Soon)
       - ToolGroups
       - ToolRuntime
       - VectorIO

pyproject.toml

@@ -7,7 +7,7 @@ required-version = ">=0.7.0"
 
 [project]
 name = "llama_stack"
-version = "0.3.0"
+version = "0.4.0.dev0"
 authors = [{ name = "Meta Llama", email = "llama-oss@meta.com" }]
 description = "Llama Stack"
 readme = "README.md"

scripts/docker.sh

@@ -215,6 +215,16 @@ build_image() {
         --build-arg "LLAMA_STACK_DIR=/workspace"
     )
 
+    # Pass UV index configuration for release branches
+    if [[ -n "${UV_EXTRA_INDEX_URL:-}" ]]; then
+        echo "Adding UV_EXTRA_INDEX_URL to docker build: $UV_EXTRA_INDEX_URL"
+        build_cmd+=(--build-arg "UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL")
+    fi
+    if [[ -n "${UV_INDEX_STRATEGY:-}" ]]; then
+        echo "Adding UV_INDEX_STRATEGY to docker build: $UV_INDEX_STRATEGY"
+        build_cmd+=(--build-arg "UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY")
+    fi
+
     if ! "${build_cmd[@]}"; then
         echo "❌ Failed to build Docker image"
         exit 1

scripts/integration-tests.sh

@@ -23,7 +23,7 @@ COLLECT_ONLY=false
 
 # Function to display usage
 usage() {
-    cat << EOF
+    cat <<EOF
 Usage: $0 [OPTIONS]
 
 Options:
@@ -62,47 +62,46 @@ EOF
 # Parse command line arguments
 while [[ $# -gt 0 ]]; do
     case $1 in
-        --stack-config)
-            STACK_CONFIG="$2"
-            shift 2
-            ;;
-        --setup)
-            TEST_SETUP="$2"
-            shift 2
-            ;;
-        --subdirs)
-            TEST_SUBDIRS="$2"
-            shift 2
-            ;;
-        --suite)
-            TEST_SUITE="$2"
-            shift 2
-            ;;
-        --inference-mode)
-            INFERENCE_MODE="$2"
-            shift 2
-            ;;
-        --pattern)
-            TEST_PATTERN="$2"
-            shift 2
-            ;;
-        --collect-only)
-            COLLECT_ONLY=true
-            shift
-            ;;
-        --help)
-            usage
-            exit 0
-            ;;
-        *)
-            echo "Unknown option: $1"
-            usage
-            exit 1
-            ;;
+    --stack-config)
+        STACK_CONFIG="$2"
+        shift 2
+        ;;
+    --setup)
+        TEST_SETUP="$2"
+        shift 2
+        ;;
+    --subdirs)
+        TEST_SUBDIRS="$2"
+        shift 2
+        ;;
+    --suite)
+        TEST_SUITE="$2"
+        shift 2
+        ;;
+    --inference-mode)
+        INFERENCE_MODE="$2"
+        shift 2
+        ;;
+    --pattern)
+        TEST_PATTERN="$2"
+        shift 2
+        ;;
+    --collect-only)
+        COLLECT_ONLY=true
+        shift
+        ;;
+    --help)
+        usage
+        exit 0
+        ;;
+    *)
+        echo "Unknown option: $1"
+        usage
+        exit 1
+        ;;
     esac
 done
 
-
 # Validate required parameters
 if [[ -z "$STACK_CONFIG" && "$COLLECT_ONLY" == false ]]; then
     echo "Error: --stack-config is required"
@@ -177,21 +176,45 @@ cd $ROOT_DIR
 # check if "llama" and "pytest" are available. this script does not use `uv run` given
 # it can be used in a pre-release environment where we have not been able to tell
 # uv about pre-release dependencies properly (yet).
-if [[ "$COLLECT_ONLY" == false ]] && ! command -v llama &> /dev/null; then
+if [[ "$COLLECT_ONLY" == false ]] && ! command -v llama &>/dev/null; then
     echo "llama could not be found, ensure llama-stack is installed"
     exit 1
 fi
 
-if ! command -v pytest &> /dev/null; then
+if ! command -v pytest &>/dev/null; then
     echo "pytest could not be found, ensure pytest is installed"
     exit 1
 fi
 
+# Helper function to find next available port
+find_available_port() {
+    local start_port=$1
+    local port=$start_port
+    for ((i=0; i<100; i++)); do
+        if ! lsof -Pi :$port -sTCP:LISTEN -t >/dev/null 2>&1; then
+            echo $port
+            return 0
+        fi
+        ((port++))
+    done
+    echo "Failed to find available port starting from $start_port" >&2
+    return 1
+}
+
 # Start Llama Stack Server if needed
 if [[ "$STACK_CONFIG" == *"server:"* && "$COLLECT_ONLY" == false ]]; then
+    # Find an available port for the server
+    LLAMA_STACK_PORT=$(find_available_port 8321)
+    if [[ $? -ne 0 ]]; then
+        echo "Error: $LLAMA_STACK_PORT"
+        exit 1
+    fi
+    export LLAMA_STACK_PORT
+    echo "Will use port: $LLAMA_STACK_PORT"
+
     stop_server() {
         echo "Stopping Llama Stack Server..."
-        pids=$(lsof -i :8321 | awk 'NR>1 {print $2}')
+        pids=$(lsof -i :$LLAMA_STACK_PORT | awk 'NR>1 {print $2}')
         if [[ -n "$pids" ]]; then
             echo "Killing Llama Stack Server processes: $pids"
             kill -9 $pids
@@ -201,42 +224,37 @@ if [[ "$STACK_CONFIG" == *"server:"* && "$COLLECT_ONLY" == false ]]; then
         echo "Llama Stack Server stopped"
     }
 
-    # check if server is already running
-    if curl -s http://localhost:8321/v1/health 2>/dev/null | grep -q "OK"; then
-        echo "Llama Stack Server is already running, skipping start"
-    else
-        echo "=== Starting Llama Stack Server ==="
-        export LLAMA_STACK_LOG_WIDTH=120
+    echo "=== Starting Llama Stack Server ==="
+    export LLAMA_STACK_LOG_WIDTH=120
 
-        # Configure telemetry collector for server mode
-        # Use a fixed port for the OTEL collector so the server can connect to it
-        COLLECTOR_PORT=4317
-        export LLAMA_STACK_TEST_COLLECTOR_PORT="${COLLECTOR_PORT}"
-        export OTEL_EXPORTER_OTLP_ENDPOINT="http://127.0.0.1:${COLLECTOR_PORT}"
-        export OTEL_EXPORTER_OTLP_PROTOCOL="http/protobuf"
-        export OTEL_BSP_SCHEDULE_DELAY="200"
-        export OTEL_BSP_EXPORT_TIMEOUT="2000"
+    # Configure telemetry collector for server mode
+    # Use a fixed port for the OTEL collector so the server can connect to it
+    COLLECTOR_PORT=4317
+    export LLAMA_STACK_TEST_COLLECTOR_PORT="${COLLECTOR_PORT}"
+    export OTEL_EXPORTER_OTLP_ENDPOINT="http://127.0.0.1:${COLLECTOR_PORT}"
+    export OTEL_EXPORTER_OTLP_PROTOCOL="http/protobuf"
+    export OTEL_BSP_SCHEDULE_DELAY="200"
+    export OTEL_BSP_EXPORT_TIMEOUT="2000"
 
-        # remove "server:" from STACK_CONFIG
-        stack_config=$(echo "$STACK_CONFIG" | sed 's/^server://')
-        nohup llama stack run $stack_config > server.log 2>&1 &
+    # remove "server:" from STACK_CONFIG
+    stack_config=$(echo "$STACK_CONFIG" | sed 's/^server://')
+    nohup llama stack run $stack_config >server.log 2>&1 &
 
-        echo "Waiting for Llama Stack Server to start..."
-        for i in {1..30}; do
-            if curl -s http://localhost:8321/v1/health 2>/dev/null | grep -q "OK"; then
-                echo "✅ Llama Stack Server started successfully"
-                break
-            fi
-            if [[ $i -eq 30 ]]; then
-                echo "❌ Llama Stack Server failed to start"
-                echo "Server logs:"
-                cat server.log
-                exit 1
-            fi
-            sleep 1
-        done
-        echo ""
-    fi
+    echo "Waiting for Llama Stack Server to start on port $LLAMA_STACK_PORT..."
+    for i in {1..30}; do
+        if curl -s http://localhost:$LLAMA_STACK_PORT/v1/health 2>/dev/null | grep -q "OK"; then
+            echo "✅ Llama Stack Server started successfully"
+            break
+        fi
+        if [[ $i -eq 30 ]]; then
+            echo "❌ Llama Stack Server failed to start"
+            echo "Server logs:"
+            cat server.log
+            exit 1
+        fi
+        sleep 1
+    done
+    echo ""
 
     trap stop_server EXIT ERR INT TERM
 fi
@@ -248,7 +266,7 @@ if [[ "$STACK_CONFIG" == *"docker:"* && "$COLLECT_ONLY" == false ]]; then
         container_name="llama-stack-test-$DISTRO"
         if docker ps -a --format '{{.Names}}' | grep -q "^${container_name}$"; then
             echo "Dumping container logs before stopping..."
-            docker logs "$container_name" > "docker-${DISTRO}-${INFERENCE_MODE}.log" 2>&1 || true
+            docker logs "$container_name" >"docker-${DISTRO}-${INFERENCE_MODE}.log" 2>&1 || true
             echo "Stopping and removing container: $container_name"
             docker stop "$container_name" 2>/dev/null || true
             docker rm "$container_name" 2>/dev/null || true
@@ -260,7 +278,14 @@ if [[ "$STACK_CONFIG" == *"docker:"* && "$COLLECT_ONLY" == false ]]; then
 
     # Extract distribution name from docker:distro format
     DISTRO=$(echo "$STACK_CONFIG" | sed 's/^docker://')
-    export LLAMA_STACK_PORT=8321
+    # Find an available port for the docker container
+    LLAMA_STACK_PORT=$(find_available_port 8321)
+    if [[ $? -ne 0 ]]; then
+        echo "Error: $LLAMA_STACK_PORT"
+        exit 1
+    fi
+    export LLAMA_STACK_PORT
+    echo "Will use port: $LLAMA_STACK_PORT"
 
     echo "=== Building Docker Image for distribution: $DISTRO ==="
     containerfile="$ROOT_DIR/containers/Containerfile"
@@ -280,6 +305,16 @@ if [[ "$STACK_CONFIG" == *"docker:"* && "$COLLECT_ONLY" == false ]]; then
         --build-arg "LLAMA_STACK_DIR=/workspace"
     )
 
+    # Pass UV index configuration for release branches
+    if [[ -n "${UV_EXTRA_INDEX_URL:-}" ]]; then
+        echo "Adding UV_EXTRA_INDEX_URL to docker build: $UV_EXTRA_INDEX_URL"
+        build_cmd+=(--build-arg "UV_EXTRA_INDEX_URL=$UV_EXTRA_INDEX_URL")
+    fi
+    if [[ -n "${UV_INDEX_STRATEGY:-}" ]]; then
+        echo "Adding UV_INDEX_STRATEGY to docker build: $UV_INDEX_STRATEGY"
+        build_cmd+=(--build-arg "UV_INDEX_STRATEGY=$UV_INDEX_STRATEGY")
+    fi
+
     if ! "${build_cmd[@]}"; then
         echo "❌ Failed to build Docker image"
         exit 1
@@ -437,17 +472,13 @@ elif [ $exit_code -eq 5 ]; then
 else
     echo "❌ Tests failed"
     echo ""
-    echo "=== Dumping last 100 lines of logs for debugging ==="
-
     # Output server or container logs based on stack config
     if [[ "$STACK_CONFIG" == *"server:"* && -f "server.log" ]]; then
-        echo "--- Last 100 lines of server.log ---"
-        tail -100 server.log
+        echo "--- Server side failures can be located inside server.log (available from artifacts on CI) ---"
     elif [[ "$STACK_CONFIG" == *"docker:"* ]]; then
         docker_log_file="docker-${DISTRO}-${INFERENCE_MODE}.log"
         if [[ -f "$docker_log_file" ]]; then
-            echo "--- Last 100 lines of $docker_log_file ---"
-            tail -100 "$docker_log_file"
+            echo "--- Server side failures can be located inside $docker_log_file (available from artifacts on CI) ---"
         fi
     fi
 

scripts/uv-run-with-index.sh

@@ -0,0 +1,42 @@
+#!/bin/bash
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+set -euo pipefail
+
+# Detect current branch and target branch
+# In GitHub Actions, use GITHUB_REF/GITHUB_BASE_REF
+if [[ -n "${GITHUB_REF:-}" ]]; then
+  BRANCH="${GITHUB_REF#refs/heads/}"
+else
+  BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null || echo "")
+fi
+
+# For PRs, check the target branch
+if [[ -n "${GITHUB_BASE_REF:-}" ]]; then
+  TARGET_BRANCH="${GITHUB_BASE_REF}"
+else
+  TARGET_BRANCH=$(git rev-parse --abbrev-ref HEAD@{upstream} 2>/dev/null | sed 's|origin/||' || echo "")
+fi
+
+# Check if on a release branch or targeting one, or LLAMA_STACK_RELEASE_MODE is set
+IS_RELEASE=false
+if [[ "$BRANCH" =~ ^release-[0-9]+\.[0-9]+\.x$ ]]; then
+  IS_RELEASE=true
+elif [[ "$TARGET_BRANCH" =~ ^release-[0-9]+\.[0-9]+\.x$ ]]; then
+  IS_RELEASE=true
+elif [[ "${LLAMA_STACK_RELEASE_MODE:-}" == "true" ]]; then
+  IS_RELEASE=true
+fi
+
+# On release branches, use test.pypi as extra index for RC versions
+if [[ "$IS_RELEASE" == "true" ]]; then
+  export UV_EXTRA_INDEX_URL="https://test.pypi.org/simple/"
+  export UV_INDEX_STRATEGY="unsafe-best-match"
+fi
+
+# Run uv with all arguments passed through
+exec uv "$@"

src/llama_stack/apis/agents/agents.py

@@ -491,13 +491,6 @@ class Agents(Protocol):
 
     APIs for creating and interacting with agentic systems."""
 
-    @webmethod(
-        route="/agents",
-        method="POST",
-        descriptive_name="create_agent",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents",
         method="POST",
@@ -515,13 +508,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}/turn",
-        method="POST",
-        descriptive_name="create_agent_turn",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}/turn",
         method="POST",
@@ -552,13 +538,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}/resume",
-        method="POST",
-        descriptive_name="resume_agent_turn",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}/resume",
         method="POST",
@@ -586,12 +565,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}",
-        method="GET",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}",
         method="GET",
@@ -612,12 +585,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}/step/{step_id}",
-        method="GET",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}/turn/{turn_id}/step/{step_id}",
         method="GET",
@@ -640,13 +607,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session",
-        method="POST",
-        descriptive_name="create_agent_session",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session",
         method="POST",
@@ -666,12 +626,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}",
-        method="GET",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}",
         method="GET",
@@ -692,12 +646,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/session/{session_id}",
-        method="DELETE",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(
         route="/agents/{agent_id}/session/{session_id}",
         method="DELETE",
@@ -715,12 +663,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}",
-        method="DELETE",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(route="/agents/{agent_id}", method="DELETE", level=LLAMA_STACK_API_V1ALPHA)
     async def delete_agent(
         self,
@@ -732,7 +674,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/agents", method="GET", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/agents", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def list_agents(self, start_index: int | None = None, limit: int | None = None) -> PaginatedResponse:
         """List all agents.
@@ -743,12 +684,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}",
-        method="GET",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(route="/agents/{agent_id}", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def get_agent(self, agent_id: str) -> Agent:
         """Describe an agent by its ID.
@@ -758,12 +693,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/agents/{agent_id}/sessions",
-        method="GET",
-        deprecated=True,
-        level=LLAMA_STACK_API_V1,
-    )
     @webmethod(route="/agents/{agent_id}/sessions", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def list_agent_sessions(
         self,
@@ -787,12 +716,6 @@ class Agents(Protocol):
     #
     # Both of these APIs are inherently stateful.
 
-    @webmethod(
-        route="/openai/v1/responses/{response_id}",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(route="/responses/{response_id}", method="GET", level=LLAMA_STACK_API_V1)
     async def get_openai_response(
         self,
@@ -805,7 +728,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/responses", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/responses", method="POST", level=LLAMA_STACK_API_V1)
     async def create_openai_response(
         self,
@@ -842,7 +764,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/responses", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/responses", method="GET", level=LLAMA_STACK_API_V1)
     async def list_openai_responses(
         self,
@@ -861,9 +782,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/responses/{response_id}/input_items", method="GET", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(route="/responses/{response_id}/input_items", method="GET", level=LLAMA_STACK_API_V1)
     async def list_openai_response_input_items(
         self,
@@ -886,7 +804,6 @@ class Agents(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/responses/{response_id}", method="DELETE", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/responses/{response_id}", method="DELETE", level=LLAMA_STACK_API_V1)
     async def delete_openai_response(self, response_id: str) -> OpenAIDeleteResponseObject:
         """Delete a response.

src/llama_stack/apis/batches/batches.py

@@ -43,7 +43,6 @@ class Batches(Protocol):
     Note: This API is currently under active development and may undergo changes.
     """
 
-    @webmethod(route="/openai/v1/batches", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/batches", method="POST", level=LLAMA_STACK_API_V1)
     async def create_batch(
         self,
@@ -64,7 +63,6 @@ class Batches(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/batches/{batch_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/batches/{batch_id}", method="GET", level=LLAMA_STACK_API_V1)
     async def retrieve_batch(self, batch_id: str) -> BatchObject:
         """Retrieve information about a specific batch.
@@ -74,7 +72,6 @@ class Batches(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/batches/{batch_id}/cancel", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/batches/{batch_id}/cancel", method="POST", level=LLAMA_STACK_API_V1)
     async def cancel_batch(self, batch_id: str) -> BatchObject:
         """Cancel a batch that is in progress.
@@ -84,7 +81,6 @@ class Batches(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/batches", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/batches", method="GET", level=LLAMA_STACK_API_V1)
     async def list_batches(
         self,

src/llama_stack/apis/benchmarks/benchmarks.py

@@ -8,7 +8,7 @@ from typing import Any, Literal, Protocol, runtime_checkable
 from pydantic import BaseModel, Field
 
 from llama_stack.apis.resource import Resource, ResourceType
-from llama_stack.apis.version import LLAMA_STACK_API_V1, LLAMA_STACK_API_V1ALPHA
+from llama_stack.apis.version import LLAMA_STACK_API_V1ALPHA
 from llama_stack.schema_utils import json_schema_type, webmethod
 
 
@@ -54,7 +54,6 @@ class ListBenchmarksResponse(BaseModel):
 
 @runtime_checkable
 class Benchmarks(Protocol):
-    @webmethod(route="/eval/benchmarks", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/eval/benchmarks", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def list_benchmarks(self) -> ListBenchmarksResponse:
         """List all benchmarks.
@@ -63,7 +62,6 @@ class Benchmarks(Protocol):
         """
         ...
 
-    @webmethod(route="/eval/benchmarks/{benchmark_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/eval/benchmarks/{benchmark_id}", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def get_benchmark(
         self,
@@ -76,7 +74,6 @@ class Benchmarks(Protocol):
         """
         ...
 
-    @webmethod(route="/eval/benchmarks", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/eval/benchmarks", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def register_benchmark(
         self,
@@ -98,7 +95,6 @@ class Benchmarks(Protocol):
         """
         ...
 
-    @webmethod(route="/eval/benchmarks/{benchmark_id}", method="DELETE", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/eval/benchmarks/{benchmark_id}", method="DELETE", level=LLAMA_STACK_API_V1ALPHA)
     async def unregister_benchmark(self, benchmark_id: str) -> None:
         """Unregister a benchmark.

src/llama_stack/apis/datasetio/datasetio.py

@@ -8,7 +8,7 @@ from typing import Any, Protocol, runtime_checkable
 
 from llama_stack.apis.common.responses import PaginatedResponse
 from llama_stack.apis.datasets import Dataset
-from llama_stack.apis.version import LLAMA_STACK_API_V1, LLAMA_STACK_API_V1BETA
+from llama_stack.apis.version import LLAMA_STACK_API_V1BETA
 from llama_stack.schema_utils import webmethod
 
 
@@ -21,7 +21,6 @@ class DatasetIO(Protocol):
     # keeping for aligning with inference/safety, but this is not used
     dataset_store: DatasetStore
 
-    @webmethod(route="/datasetio/iterrows/{dataset_id:path}", method="GET", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/datasetio/iterrows/{dataset_id:path}", method="GET", level=LLAMA_STACK_API_V1BETA)
     async def iterrows(
         self,
@@ -46,9 +45,6 @@ class DatasetIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/datasetio/append-rows/{dataset_id:path}", method="POST", deprecated=True, level=LLAMA_STACK_API_V1
-    )
     @webmethod(route="/datasetio/append-rows/{dataset_id:path}", method="POST", level=LLAMA_STACK_API_V1BETA)
     async def append_rows(self, dataset_id: str, rows: list[dict[str, Any]]) -> None:
         """Append rows to a dataset.

src/llama_stack/apis/datasets/datasets.py

@@ -10,7 +10,7 @@ from typing import Annotated, Any, Literal, Protocol
 from pydantic import BaseModel, Field
 
 from llama_stack.apis.resource import Resource, ResourceType
-from llama_stack.apis.version import LLAMA_STACK_API_V1, LLAMA_STACK_API_V1BETA
+from llama_stack.apis.version import LLAMA_STACK_API_V1BETA
 from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
 
 
@@ -146,7 +146,6 @@ class ListDatasetsResponse(BaseModel):
 
 
 class Datasets(Protocol):
-    @webmethod(route="/datasets", method="POST", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/datasets", method="POST", level=LLAMA_STACK_API_V1BETA)
     async def register_dataset(
         self,
@@ -216,7 +215,6 @@ class Datasets(Protocol):
         """
         ...
 
-    @webmethod(route="/datasets/{dataset_id:path}", method="GET", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/datasets/{dataset_id:path}", method="GET", level=LLAMA_STACK_API_V1BETA)
     async def get_dataset(
         self,
@@ -229,7 +227,6 @@ class Datasets(Protocol):
         """
         ...
 
-    @webmethod(route="/datasets", method="GET", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/datasets", method="GET", level=LLAMA_STACK_API_V1BETA)
     async def list_datasets(self) -> ListDatasetsResponse:
         """List all datasets.
@@ -238,7 +235,6 @@ class Datasets(Protocol):
         """
         ...
 
-    @webmethod(route="/datasets/{dataset_id:path}", method="DELETE", deprecated=True, level=LLAMA_STACK_API_V1)
     @webmethod(route="/datasets/{dataset_id:path}", method="DELETE", level=LLAMA_STACK_API_V1BETA)
     async def unregister_dataset(
         self,

src/llama_stack/apis/eval/eval.py

@@ -13,7 +13,7 @@ from llama_stack.apis.common.job_types import Job
 from llama_stack.apis.inference import SamplingParams, SystemMessage
 from llama_stack.apis.scoring import ScoringResult
 from llama_stack.apis.scoring_functions import ScoringFnParams
-from llama_stack.apis.version import LLAMA_STACK_API_V1, LLAMA_STACK_API_V1ALPHA
+from llama_stack.apis.version import LLAMA_STACK_API_V1ALPHA
 from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
 
 
@@ -86,7 +86,6 @@ class Eval(Protocol):
 
     Llama Stack Evaluation API for running evaluations on model and agent candidates."""
 
-    @webmethod(route="/eval/benchmarks/{benchmark_id}/jobs", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/eval/benchmarks/{benchmark_id}/jobs", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def run_eval(
         self,
@@ -101,9 +100,6 @@ class Eval(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/eval/benchmarks/{benchmark_id}/evaluations", method="POST", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(route="/eval/benchmarks/{benchmark_id}/evaluations", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def evaluate_rows(
         self,
@@ -122,9 +118,6 @@ class Eval(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def job_status(self, benchmark_id: str, job_id: str) -> Job:
         """Get the status of a job.
@@ -135,12 +128,6 @@ class Eval(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}",
-        method="DELETE",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="DELETE", level=LLAMA_STACK_API_V1ALPHA)
     async def job_cancel(self, benchmark_id: str, job_id: str) -> None:
         """Cancel a job.
@@ -150,12 +137,6 @@ class Eval(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}/result",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}/result", method="GET", level=LLAMA_STACK_API_V1ALPHA
     )

src/llama_stack/apis/files/files.py

@@ -110,7 +110,6 @@ class Files(Protocol):
     """
 
     # OpenAI Files API Endpoints
-    @webmethod(route="/openai/v1/files", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/files", method="POST", level=LLAMA_STACK_API_V1)
     async def openai_upload_file(
         self,
@@ -134,7 +133,6 @@ class Files(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/files", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/files", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_list_files(
         self,
@@ -155,7 +153,6 @@ class Files(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/files/{file_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/files/{file_id}", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_retrieve_file(
         self,
@@ -170,7 +167,6 @@ class Files(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/files/{file_id}", method="DELETE", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/files/{file_id}", method="DELETE", level=LLAMA_STACK_API_V1)
     async def openai_delete_file(
         self,
@@ -183,7 +179,6 @@ class Files(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/files/{file_id}/content", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/files/{file_id}/content", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_retrieve_file_content(
         self,

src/llama_stack/apis/inference/inference.py

@@ -1189,7 +1189,6 @@ class InferenceProvider(Protocol):
         raise NotImplementedError("Reranking is not implemented")
         return  # this is so mypy's safe-super rule will consider the method concrete
 
-    @webmethod(route="/openai/v1/completions", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/completions", method="POST", level=LLAMA_STACK_API_V1)
     async def openai_completion(
         self,
@@ -1202,7 +1201,6 @@ class InferenceProvider(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/chat/completions", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/chat/completions", method="POST", level=LLAMA_STACK_API_V1)
     async def openai_chat_completion(
         self,
@@ -1215,7 +1213,6 @@ class InferenceProvider(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/embeddings", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/embeddings", method="POST", level=LLAMA_STACK_API_V1)
     async def openai_embeddings(
         self,
@@ -1240,7 +1237,6 @@ class Inference(InferenceProvider):
     - Rerank models: these models reorder the documents based on their relevance to a query.
     """
 
-    @webmethod(route="/openai/v1/chat/completions", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/chat/completions", method="GET", level=LLAMA_STACK_API_V1)
     async def list_chat_completions(
         self,
@@ -1259,9 +1255,6 @@ class Inference(InferenceProvider):
         """
         raise NotImplementedError("List chat completions is not implemented")
 
-    @webmethod(
-        route="/openai/v1/chat/completions/{completion_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(route="/chat/completions/{completion_id}", method="GET", level=LLAMA_STACK_API_V1)
     async def get_chat_completion(self, completion_id: str) -> OpenAICompletionWithInputMessages:
         """Get chat completion.

src/llama_stack/apis/inspect/inspect.py

@@ -4,14 +4,21 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
-from typing import Protocol, runtime_checkable
+from typing import Literal, Protocol, runtime_checkable
 
 from pydantic import BaseModel
 
-from llama_stack.apis.version import LLAMA_STACK_API_V1
+from llama_stack.apis.version import (
+    LLAMA_STACK_API_V1,
+)
 from llama_stack.providers.datatypes import HealthStatus
 from llama_stack.schema_utils import json_schema_type, webmethod
 
+# Valid values for the route filter parameter.
+# Actual API levels: v1, v1alpha, v1beta (filters by level, excludes deprecated)
+# Special filter value: "deprecated" (shows deprecated routes regardless of level)
+ApiFilter = Literal["v1", "v1alpha", "v1beta", "deprecated"]
+
 
 @json_schema_type
 class RouteInfo(BaseModel):
@@ -64,11 +71,12 @@ class Inspect(Protocol):
     """
 
     @webmethod(route="/inspect/routes", method="GET", level=LLAMA_STACK_API_V1)
-    async def list_routes(self) -> ListRoutesResponse:
+    async def list_routes(self, api_filter: ApiFilter | None = None) -> ListRoutesResponse:
         """List routes.
 
         List all available API routes with their methods and implementing providers.
 
+        :param api_filter: Optional filter to control which routes are returned. Can be an API level ('v1', 'v1alpha', 'v1beta') to show non-deprecated routes at that level, or 'deprecated' to show deprecated routes across all levels. If not specified, returns only non-deprecated v1 routes.
         :returns: Response containing information about all available routes.
         """
         ...

src/llama_stack/apis/models/models.py

@@ -90,12 +90,14 @@ class OpenAIModel(BaseModel):
     :object: The object type, which will be "model"
     :created: The Unix timestamp in seconds when the model was created
     :owned_by: The owner of the model
+    :custom_metadata: Llama Stack-specific metadata including model_type, provider info, and additional metadata
     """
 
     id: str
     object: Literal["model"] = "model"
     created: int
     owned_by: str
+    custom_metadata: dict[str, Any] | None = None
 
 
 class OpenAIListModelsResponse(BaseModel):
@@ -105,7 +107,6 @@ class OpenAIListModelsResponse(BaseModel):
 @runtime_checkable
 @trace_protocol
 class Models(Protocol):
-    @webmethod(route="/models", method="GET", level=LLAMA_STACK_API_V1)
     async def list_models(self) -> ListModelsResponse:
         """List all models.
 
@@ -113,7 +114,7 @@ class Models(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/models", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
+    @webmethod(route="/models", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_list_models(self) -> OpenAIListModelsResponse:
         """List models using the OpenAI API.
 

src/llama_stack/apis/post_training/post_training.py

@@ -13,7 +13,7 @@ from pydantic import BaseModel, Field
 from llama_stack.apis.common.content_types import URL
 from llama_stack.apis.common.job_types import JobStatus
 from llama_stack.apis.common.training_types import Checkpoint
-from llama_stack.apis.version import LLAMA_STACK_API_V1, LLAMA_STACK_API_V1ALPHA
+from llama_stack.apis.version import LLAMA_STACK_API_V1ALPHA
 from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
 
 
@@ -284,7 +284,6 @@ class PostTrainingJobArtifactsResponse(BaseModel):
 
 
 class PostTraining(Protocol):
-    @webmethod(route="/post-training/supervised-fine-tune", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/supervised-fine-tune", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def supervised_fine_tune(
         self,
@@ -312,7 +311,6 @@ class PostTraining(Protocol):
         """
         ...
 
-    @webmethod(route="/post-training/preference-optimize", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/preference-optimize", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def preference_optimize(
         self,
@@ -335,7 +333,6 @@ class PostTraining(Protocol):
         """
         ...
 
-    @webmethod(route="/post-training/jobs", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/jobs", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def get_training_jobs(self) -> ListPostTrainingJobsResponse:
         """Get all training jobs.
@@ -344,7 +341,6 @@ class PostTraining(Protocol):
         """
         ...
 
-    @webmethod(route="/post-training/job/status", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/job/status", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def get_training_job_status(self, job_uuid: str) -> PostTrainingJobStatusResponse:
         """Get the status of a training job.
@@ -354,7 +350,6 @@ class PostTraining(Protocol):
         """
         ...
 
-    @webmethod(route="/post-training/job/cancel", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/job/cancel", method="POST", level=LLAMA_STACK_API_V1ALPHA)
     async def cancel_training_job(self, job_uuid: str) -> None:
         """Cancel a training job.
@@ -363,7 +358,6 @@ class PostTraining(Protocol):
         """
         ...
 
-    @webmethod(route="/post-training/job/artifacts", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/post-training/job/artifacts", method="GET", level=LLAMA_STACK_API_V1ALPHA)
     async def get_training_job_artifacts(self, job_uuid: str) -> PostTrainingJobArtifactsResponse:
         """Get the artifacts of a training job.

src/llama_stack/apis/safety/safety.py

@@ -121,7 +121,6 @@ class Safety(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/moderations", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/moderations", method="POST", level=LLAMA_STACK_API_V1)
     async def run_moderation(self, input: str | list[str], model: str | None = None) -> ModerationObject:
         """Create moderation.

src/llama_stack/apis/synthetic_data_generation/__init__.py

@@ -1,7 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-from .synthetic_data_generation import *

src/llama_stack/apis/synthetic_data_generation/synthetic_data_generation.py

@@ -1,77 +0,0 @@
-# Copyright (c) Meta Platforms, Inc. and affiliates.
-# All rights reserved.
-#
-# This source code is licensed under the terms described in the LICENSE file in
-# the root directory of this source tree.
-
-from enum import Enum
-from typing import Any, Protocol
-
-from pydantic import BaseModel
-
-from llama_stack.apis.inference import Message
-from llama_stack.apis.version import LLAMA_STACK_API_V1
-from llama_stack.schema_utils import json_schema_type, webmethod
-
-
-class FilteringFunction(Enum):
-    """The type of filtering function.
-
-    :cvar none: No filtering applied, accept all generated synthetic data
-    :cvar random: Random sampling of generated data points
-    :cvar top_k: Keep only the top-k highest scoring synthetic data samples
-    :cvar top_p: Nucleus-style filtering, keep samples exceeding cumulative score threshold
-    :cvar top_k_top_p: Combined top-k and top-p filtering strategy
-    :cvar sigmoid: Apply sigmoid function for probability-based filtering
-    """
-
-    none = "none"
-    random = "random"
-    top_k = "top_k"
-    top_p = "top_p"
-    top_k_top_p = "top_k_top_p"
-    sigmoid = "sigmoid"
-
-
-@json_schema_type
-class SyntheticDataGenerationRequest(BaseModel):
-    """Request to generate synthetic data. A small batch of prompts and a filtering function
-
-    :param dialogs: List of conversation messages to use as input for synthetic data generation
-    :param filtering_function: Type of filtering to apply to generated synthetic data samples
-    :param model: (Optional) The identifier of the model to use. The model must be registered with Llama Stack and available via the /models endpoint
-    """
-
-    dialogs: list[Message]
-    filtering_function: FilteringFunction = FilteringFunction.none
-    model: str | None = None
-
-
-@json_schema_type
-class SyntheticDataGenerationResponse(BaseModel):
-    """Response from the synthetic data generation. Batch of (prompt, response, score) tuples that pass the threshold.
-
-    :param synthetic_data: List of generated synthetic data samples that passed the filtering criteria
-    :param statistics: (Optional) Statistical information about the generation process and filtering results
-    """
-
-    synthetic_data: list[dict[str, Any]]
-    statistics: dict[str, Any] | None = None
-
-
-class SyntheticDataGeneration(Protocol):
-    @webmethod(route="/synthetic-data-generation/generate", level=LLAMA_STACK_API_V1)
-    def synthetic_data_generate(
-        self,
-        dialogs: list[Message],
-        filtering_function: FilteringFunction = FilteringFunction.none,
-        model: str | None = None,
-    ) -> SyntheticDataGenerationResponse:
-        """Generate synthetic data based on input dialogs and apply filtering.
-
-        :param dialogs: List of conversation messages to use as input for synthetic data generation
-        :param filtering_function: Type of filtering to apply to generated synthetic data samples
-        :param model: (Optional) The identifier of the model to use. The model must be registered with Llama Stack and available via the /models endpoint
-        :returns: Response containing filtered synthetic data samples and optional statistics
-        """
-        ...

src/llama_stack/apis/vector_io/vector_io.py

@@ -545,7 +545,6 @@ class VectorIO(Protocol):
         ...
 
     # OpenAI Vector Stores API endpoints
-    @webmethod(route="/openai/v1/vector_stores", method="POST", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/vector_stores", method="POST", level=LLAMA_STACK_API_V1)
     async def openai_create_vector_store(
         self,
@@ -558,7 +557,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(route="/openai/v1/vector_stores", method="GET", level=LLAMA_STACK_API_V1, deprecated=True)
     @webmethod(route="/vector_stores", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_list_vector_stores(
         self,
@@ -577,9 +575,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}", method="GET", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(route="/vector_stores/{vector_store_id}", method="GET", level=LLAMA_STACK_API_V1)
     async def openai_retrieve_vector_store(
         self,
@@ -592,9 +587,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}", method="POST", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}",
         method="POST",
@@ -617,9 +609,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}", method="DELETE", level=LLAMA_STACK_API_V1, deprecated=True
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}",
         method="DELETE",
@@ -636,12 +625,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/search",
-        method="POST",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/search",
         method="POST",
@@ -674,12 +657,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files",
-        method="POST",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files",
         method="POST",
@@ -702,12 +679,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files",
         method="GET",
@@ -734,12 +705,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files/{file_id}",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files/{file_id}",
         method="GET",
@@ -758,12 +723,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files/{file_id}/content",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files/{file_id}/content",
         method="GET",
@@ -782,12 +741,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files/{file_id}",
-        method="POST",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files/{file_id}",
         method="POST",
@@ -808,12 +761,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/files/{file_id}",
-        method="DELETE",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/files/{file_id}",
         method="DELETE",
@@ -837,12 +784,6 @@ class VectorIO(Protocol):
         method="POST",
         level=LLAMA_STACK_API_V1,
     )
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/file_batches",
-        method="POST",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     async def openai_create_vector_store_file_batch(
         self,
         vector_store_id: str,
@@ -861,12 +802,6 @@ class VectorIO(Protocol):
         method="GET",
         level=LLAMA_STACK_API_V1,
     )
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     async def openai_retrieve_vector_store_file_batch(
         self,
         batch_id: str,
@@ -880,12 +815,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/files",
-        method="GET",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/file_batches/{batch_id}/files",
         method="GET",
@@ -914,12 +843,6 @@ class VectorIO(Protocol):
         """
         ...
 
-    @webmethod(
-        route="/openai/v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/cancel",
-        method="POST",
-        level=LLAMA_STACK_API_V1,
-        deprecated=True,
-    )
     @webmethod(
         route="/vector_stores/{vector_store_id}/file_batches/{batch_id}/cancel",
         method="POST",

src/llama_stack/cli/stack/run.py

@@ -8,16 +8,30 @@ import argparse
 import os
 import ssl
 import subprocess
+import sys
 from pathlib import Path
 
 import uvicorn
 import yaml
+from termcolor import cprint
 
 from llama_stack.cli.stack.utils import ImageType
 from llama_stack.cli.subcommand import Subcommand
-from llama_stack.core.datatypes import StackRunConfig
+from llama_stack.core.datatypes import Api, Provider, StackRunConfig
+from llama_stack.core.distribution import get_provider_registry
 from llama_stack.core.stack import cast_image_name_to_string, replace_env_vars
+from llama_stack.core.storage.datatypes import (
+    InferenceStoreReference,
+    KVStoreReference,
+    ServerStoresConfig,
+    SqliteKVStoreConfig,
+    SqliteSqlStoreConfig,
+    SqlStoreReference,
+    StorageConfig,
+)
+from llama_stack.core.utils.config_dirs import DISTRIBS_BASE_DIR
 from llama_stack.core.utils.config_resolution import Mode, resolve_config_or_distro
+from llama_stack.core.utils.dynamic import instantiate_class_type
 from llama_stack.log import LoggingConfig, get_logger
 
 REPO_ROOT = Path(__file__).parent.parent.parent.parent
@@ -68,6 +82,12 @@ class StackRun(Subcommand):
             action="store_true",
             help="Start the UI server",
         )
+        self.parser.add_argument(
+            "--providers",
+            type=str,
+            default=None,
+            help="Run a stack with only a list of providers. This list is formatted like: api1=provider1,api1=provider2,api2=provider3. Where there can be multiple providers per API.",
+        )
 
     def _run_stack_run_cmd(self, args: argparse.Namespace) -> None:
         import yaml
@@ -93,6 +113,55 @@ class StackRun(Subcommand):
                 config_file = resolve_config_or_distro(args.config, Mode.RUN)
             except ValueError as e:
                 self.parser.error(str(e))
+        elif args.providers:
+            provider_list: dict[str, list[Provider]] = dict()
+            for api_provider in args.providers.split(","):
+                if "=" not in api_provider:
+                    cprint(
+                        "Could not parse `--providers`. Please ensure the list is in the format api1=provider1,api2=provider2",
+                        color="red",
+                        file=sys.stderr,
+                    )
+                    sys.exit(1)
+                api, provider_type = api_provider.split("=")
+                providers_for_api = get_provider_registry().get(Api(api), None)
+                if providers_for_api is None:
+                    cprint(
+                        f"{api} is not a valid API.",
+                        color="red",
+                        file=sys.stderr,
+                    )
+                    sys.exit(1)
+                if provider_type in providers_for_api:
+                    config_type = instantiate_class_type(providers_for_api[provider_type].config_class)
+                    if config_type is not None and hasattr(config_type, "sample_run_config"):
+                        config = config_type.sample_run_config(__distro_dir__="~/.llama/distributions/providers-run")
+                    else:
+                        config = {}
+                    provider = Provider(
+                        provider_type=provider_type,
+                        config=config,
+                        provider_id=provider_type.split("::")[1],
+                    )
+                    provider_list.setdefault(api, []).append(provider)
+                else:
+                    cprint(
+                        f"{provider} is not a valid provider for the {api} API.",
+                        color="red",
+                        file=sys.stderr,
+                    )
+                    sys.exit(1)
+            run_config = self._generate_run_config_from_providers(providers=provider_list)
+            config_dict = run_config.model_dump(mode="json")
+
+            # Write config to disk in providers-run directory
+            distro_dir = DISTRIBS_BASE_DIR / "providers-run"
+            config_file = distro_dir / "run.yaml"
+
+            logger.info(f"Writing generated config to: {config_file}")
+            with open(config_file, "w") as f:
+                yaml.dump(config_dict, f, default_flow_style=False, sort_keys=False)
+
         else:
             config_file = None
 
@@ -214,3 +283,44 @@ class StackRun(Subcommand):
             )
         except Exception as e:
             logger.error(f"Failed to start UI development server in {ui_dir}: {e}")
+
+    def _generate_run_config_from_providers(self, providers: dict[str, list[Provider]]):
+        apis = list(providers.keys())
+        distro_dir = DISTRIBS_BASE_DIR / "providers-run"
+        # need somewhere to put the storage.
+        os.makedirs(distro_dir, exist_ok=True)
+        storage = StorageConfig(
+            backends={
+                "kv_default": SqliteKVStoreConfig(
+                    db_path=f"${{env.SQLITE_STORE_DIR:={distro_dir}}}/kvstore.db",
+                ),
+                "sql_default": SqliteSqlStoreConfig(
+                    db_path=f"${{env.SQLITE_STORE_DIR:={distro_dir}}}/sql_store.db",
+                ),
+            },
+            stores=ServerStoresConfig(
+                metadata=KVStoreReference(
+                    backend="kv_default",
+                    namespace="registry",
+                ),
+                inference=InferenceStoreReference(
+                    backend="sql_default",
+                    table_name="inference_store",
+                ),
+                conversations=SqlStoreReference(
+                    backend="sql_default",
+                    table_name="openai_conversations",
+                ),
+                prompts=KVStoreReference(
+                    backend="kv_default",
+                    namespace="prompts",
+                ),
+            ),
+        )
+
+        return StackRunConfig(
+            image_name="providers-run",
+            apis=apis,
+            providers=providers,
+            storage=storage,
+        )

src/llama_stack/core/inspect.py

@@ -15,6 +15,7 @@ from llama_stack.apis.inspect import (
     RouteInfo,
     VersionInfo,
 )
+from llama_stack.apis.version import LLAMA_STACK_API_V1
 from llama_stack.core.datatypes import StackRunConfig
 from llama_stack.core.external import load_external_apis
 from llama_stack.core.server.routes import get_all_api_routes
@@ -39,9 +40,21 @@ class DistributionInspectImpl(Inspect):
     async def initialize(self) -> None:
         pass
 
-    async def list_routes(self) -> ListRoutesResponse:
+    async def list_routes(self, api_filter: str | None = None) -> ListRoutesResponse:
         run_config: StackRunConfig = self.config.run_config
 
+        # Helper function to determine if a route should be included based on api_filter
+        def should_include_route(webmethod) -> bool:
+            if api_filter is None:
+                # Default: only non-deprecated v1 APIs
+                return not webmethod.deprecated and webmethod.level == LLAMA_STACK_API_V1
+            elif api_filter == "deprecated":
+                # Special filter: show deprecated routes regardless of their actual level
+                return bool(webmethod.deprecated)
+            else:
+                # Filter by API level (non-deprecated routes only)
+                return not webmethod.deprecated and webmethod.level == api_filter
+
         ret = []
         external_apis = load_external_apis(run_config)
         all_endpoints = get_all_api_routes(external_apis)
@@ -55,8 +68,8 @@ class DistributionInspectImpl(Inspect):
                             method=next(iter([m for m in e.methods if m != "HEAD"])),
                             provider_types=[],  # These APIs don't have "real" providers - they're internal to the stack
                         )
-                        for e, _ in endpoints
-                        if e.methods is not None
+                        for e, webmethod in endpoints
+                        if e.methods is not None and should_include_route(webmethod)
                     ]
                 )
             else:
@@ -69,8 +82,8 @@ class DistributionInspectImpl(Inspect):
                                 method=next(iter([m for m in e.methods if m != "HEAD"])),
                                 provider_types=[p.provider_type for p in providers],
                             )
-                            for e, _ in endpoints
-                            if e.methods is not None
+                            for e, webmethod in endpoints
+                            if e.methods is not None and should_include_route(webmethod)
                         ]
                     )
 

src/llama_stack/core/routing_tables/models.py

@@ -134,6 +134,12 @@ class ModelsRoutingTable(CommonRoutingTableImpl, Models):
                 object="model",
                 created=int(time.time()),
                 owned_by="llama_stack",
+                custom_metadata={
+                    "model_type": model.model_type,
+                    "provider_id": model.provider_id,
+                    "provider_resource_id": model.provider_resource_id,
+                    **model.metadata,
+                },
             )
             for model in all_models
         ]

src/llama_stack/core/stack.py

@@ -31,7 +31,6 @@ from llama_stack.apis.safety import Safety
 from llama_stack.apis.scoring import Scoring
 from llama_stack.apis.scoring_functions import ScoringFunctions
 from llama_stack.apis.shields import Shields
-from llama_stack.apis.synthetic_data_generation import SyntheticDataGeneration
 from llama_stack.apis.tools import RAGToolRuntime, ToolGroups, ToolRuntime
 from llama_stack.apis.vector_io import VectorIO
 from llama_stack.core.conversations.conversations import ConversationServiceConfig, ConversationServiceImpl
@@ -66,7 +65,6 @@ class LlamaStack(
     Agents,
     Batches,
     Safety,
-    SyntheticDataGeneration,
     Datasets,
     PostTraining,
     VectorIO,

src/llama_stack/core/ui/page/distribution/models.py

@@ -12,7 +12,7 @@ from llama_stack.core.ui.modules.api import llama_stack_api
 def models():
     # Models Section
     st.header("Models")
-    models_info = {m.identifier: m.to_dict() for m in llama_stack_api.client.models.list()}
+    models_info = {m.id: m.model_dump() for m in llama_stack_api.client.models.list()}
 
     selected_model = st.selectbox("Select a model", list(models_info.keys()))
     st.json(models_info[selected_model])

src/llama_stack/core/ui/page/playground/chat.py

@@ -12,7 +12,11 @@ from llama_stack.core.ui.modules.api import llama_stack_api
 with st.sidebar:
     st.header("Configuration")
     available_models = llama_stack_api.client.models.list()
-    available_models = [model.identifier for model in available_models if model.model_type == "llm"]
+    available_models = [
+        model.id
+        for model in available_models
+        if model.custom_metadata and model.custom_metadata.get("model_type") == "llm"
+    ]
     selected_model = st.selectbox(
         "Choose a model",
         available_models,

src/llama_stack/providers/inline/agents/meta_reference/agent_instance.py

@@ -1015,7 +1015,7 @@ async def load_data_from_url(url: str) -> str:
     if url.startswith("http"):
         async with httpx.AsyncClient() as client:
             r = await client.get(url)
-            resp = r.text
+            resp: str = r.text
             return resp
     raise ValueError(f"Unexpected URL: {type(url)}")
 

src/llama_stack/providers/remote/inference/vertexai/vertexai.py

@@ -4,6 +4,7 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
+from collections.abc import Iterable
 
 import google.auth.transport.requests
 from google.auth import default
@@ -42,3 +43,12 @@ class VertexAIInferenceAdapter(OpenAIMixin):
         Source: https://cloud.google.com/vertex-ai/generative-ai/docs/start/openai
         """
         return f"https://{self.config.location}-aiplatform.googleapis.com/v1/projects/{self.config.project}/locations/{self.config.location}/endpoints/openapi"
+
+    async def list_provider_model_ids(self) -> Iterable[str]:
+        """
+        VertexAI doesn't currently offer a way to query a list of available models from Google's Model Garden
+        For now we return a hardcoded version of the available models
+
+        :return: An iterable of model IDs
+        """
+        return ["vertexai/gemini-2.0-flash", "vertexai/gemini-2.5-flash", "vertexai/gemini-2.5-pro"]

src/llama_stack/providers/utils/inference/inference_store.py

@@ -35,6 +35,7 @@ class InferenceStore:
         self.reference = reference
         self.sql_store = None
         self.policy = policy
+        self.enable_write_queue = True
 
         # Async write queue and worker control
         self._queue: asyncio.Queue[tuple[OpenAIChatCompletion, list[OpenAIMessageParam]]] | None = None
@@ -47,14 +48,13 @@ class InferenceStore:
         base_store = sqlstore_impl(self.reference)
         self.sql_store = AuthorizedSqlStore(base_store, self.policy)
 
-        # Disable write queue for SQLite to avoid concurrency issues
-        backend_name = self.reference.backend
-        backend_config = _SQLSTORE_BACKENDS.get(backend_name)
-        if backend_config is None:
-            raise ValueError(
-                f"Unregistered SQL backend '{backend_name}'. Registered backends: {sorted(_SQLSTORE_BACKENDS)}"
-            )
-        self.enable_write_queue = backend_config.type != StorageBackendType.SQL_SQLITE
+        # Disable write queue for SQLite since WAL mode handles concurrency
+        # Keep it enabled for other backends (like Postgres) for performance
+        backend_config = _SQLSTORE_BACKENDS.get(self.reference.backend)
+        if backend_config and backend_config.type == StorageBackendType.SQL_SQLITE:
+            self.enable_write_queue = False
+            logger.debug("Write queue disabled for SQLite (WAL mode handles concurrency)")
+
         await self.sql_store.create_table(
             "chat_completions",
             {
@@ -70,8 +70,9 @@ class InferenceStore:
             self._queue = asyncio.Queue(maxsize=self._max_write_queue_size)
             for _ in range(self._num_writers):
                 self._worker_tasks.append(asyncio.create_task(self._worker_loop()))
-        else:
-            logger.info("Write queue disabled for SQLite to avoid concurrency issues")
+            logger.debug(
+                f"Inference store write queue enabled with {self._num_writers} writers, max queue size {self._max_write_queue_size}"
+            )
 
     async def shutdown(self) -> None:
         if not self._worker_tasks:

src/llama_stack/providers/utils/inference/model_registry.py

@@ -20,7 +20,7 @@ logger = get_logger(name=__name__, category="providers::utils")
 
 
 class RemoteInferenceProviderConfig(BaseModel):
-    allowed_models: list[str] | None = Field(  # TODO: make this non-optional and give a list() default
+    allowed_models: list[str] | None = Field(
         default=None,
         description="List of models that should be registered with the model registry. If None, all models are allowed.",
     )

src/llama_stack/providers/utils/inference/openai_mixin.py

@@ -83,9 +83,6 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
     # This is set in list_models() and used in check_model_availability()
     _model_cache: dict[str, Model] = {}
 
-    # List of allowed models for this provider, if empty all models allowed
-    allowed_models: list[str] = []
-
     # Optional field name in provider data to look for API key, which takes precedence
     provider_data_api_key_field: str | None = None
 
@@ -441,7 +438,7 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         for provider_model_id in provider_models_ids:
             if not isinstance(provider_model_id, str):
                 raise ValueError(f"Model ID {provider_model_id} from list_provider_model_ids() is not a string")
-            if self.allowed_models and provider_model_id not in self.allowed_models:
+            if self.config.allowed_models is not None and provider_model_id not in self.config.allowed_models:
                 logger.info(f"Skipping model {provider_model_id} as it is not in the allowed models list")
                 continue
             model = self.construct_model_from_identifier(provider_model_id)

src/llama_stack/providers/utils/responses/responses_store.py

@@ -70,13 +70,13 @@ class ResponsesStore:
         base_store = sqlstore_impl(self.reference)
         self.sql_store = AuthorizedSqlStore(base_store, self.policy)
 
+        # Disable write queue for SQLite since WAL mode handles concurrency
+        # Keep it enabled for other backends (like Postgres) for performance
         backend_config = _SQLSTORE_BACKENDS.get(self.reference.backend)
-        if backend_config is None:
-            raise ValueError(
-                f"Unregistered SQL backend '{self.reference.backend}'. Registered backends: {sorted(_SQLSTORE_BACKENDS)}"
-            )
-        if backend_config.type == StorageBackendType.SQL_SQLITE:
+        if backend_config and backend_config.type == StorageBackendType.SQL_SQLITE:
             self.enable_write_queue = False
+            logger.debug("Write queue disabled for SQLite (WAL mode handles concurrency)")
+
         await self.sql_store.create_table(
             "openai_responses",
             {
@@ -99,8 +99,9 @@ class ResponsesStore:
             self._queue = asyncio.Queue(maxsize=self._max_write_queue_size)
             for _ in range(self._num_writers):
                 self._worker_tasks.append(asyncio.create_task(self._worker_loop()))
-        else:
-            logger.debug("Write queue disabled for SQLite to avoid concurrency issues")
+            logger.debug(
+                f"Responses store write queue enabled with {self._num_writers} writers, max queue size {self._max_write_queue_size}"
+            )
 
     async def shutdown(self) -> None:
         if not self._worker_tasks:

src/llama_stack/providers/utils/sqlstore/sqlalchemy_sqlstore.py

@@ -17,6 +17,7 @@ from sqlalchemy import (
     String,
     Table,
     Text,
+    event,
     inspect,
     select,
     text,
@@ -75,7 +76,36 @@ class SqlAlchemySqlStoreImpl(SqlStore):
         self.metadata = MetaData()
 
     def create_engine(self) -> AsyncEngine:
-        return create_async_engine(self.config.engine_str, pool_pre_ping=True)
+        # Configure connection args for better concurrency support
+        connect_args = {}
+        if "sqlite" in self.config.engine_str:
+            # SQLite-specific optimizations for concurrent access
+            # With WAL mode, most locks resolve in milliseconds, but allow up to 5s for edge cases
+            connect_args["timeout"] = 5.0
+            connect_args["check_same_thread"] = False  # Allow usage across asyncio tasks
+
+        engine = create_async_engine(
+            self.config.engine_str,
+            pool_pre_ping=True,
+            connect_args=connect_args,
+        )
+
+        # Enable WAL mode for SQLite to support concurrent readers and writers
+        if "sqlite" in self.config.engine_str:
+
+            @event.listens_for(engine.sync_engine, "connect")
+            def set_sqlite_pragma(dbapi_conn, connection_record):
+                cursor = dbapi_conn.cursor()
+                # Enable Write-Ahead Logging for better concurrency
+                cursor.execute("PRAGMA journal_mode=WAL")
+                # Set busy timeout to 5 seconds (retry instead of immediate failure)
+                # With WAL mode, locks should be brief; if we hit 5s there's a bigger issue
+                cursor.execute("PRAGMA busy_timeout=5000")
+                # Use NORMAL synchronous mode for better performance (still safe with WAL)
+                cursor.execute("PRAGMA synchronous=NORMAL")
+                cursor.close()
+
+        return engine
 
     async def create_table(
         self,

src/llama_stack/testing/api_recorder.py

@@ -156,7 +156,7 @@ def normalize_inference_request(method: str, url: str, headers: dict[str, Any],
     }
 
     # Include test_id for isolation, except for shared infrastructure endpoints
-    if parsed.path not in ("/api/tags", "/v1/models"):
+    if parsed.path not in ("/api/tags", "/v1/models", "/v1/openai/v1/models"):
         normalized["test_id"] = test_id
 
     normalized_json = json.dumps(normalized, sort_keys=True)
@@ -430,7 +430,7 @@ class ResponseStorage:
 
         # For model-list endpoints, include digest in filename to distinguish different model sets
         endpoint = request.get("endpoint")
-        if endpoint in ("/api/tags", "/v1/models"):
+        if endpoint in ("/api/tags", "/v1/models", "/v1/openai/v1/models"):
             digest = _model_identifiers_digest(endpoint, response)
             response_file = f"models-{request_hash}-{digest}.json"
 
@@ -554,13 +554,14 @@ def _model_identifiers_digest(endpoint: str, response: dict[str, Any]) -> str:
         Supported endpoints:
         - '/api/tags' (Ollama): response body has 'models': [ { name/model/digest/id/... }, ... ]
         - '/v1/models' (OpenAI): response body is: [ { id: ... }, ... ]
+        - '/v1/openai/v1/models' (OpenAI): response body is: [ { id: ... }, ... ]
         Returns a list of unique identifiers or None if structure doesn't match.
         """
         if "models" in response["body"]:
             # ollama
             items = response["body"]["models"]
         else:
-            # openai
+            # openai or openai-style endpoints
             items = response["body"]
         idents = [m.model if endpoint == "/api/tags" else m.id for m in items]
         return sorted(set(idents))
@@ -581,7 +582,7 @@ def _combine_model_list_responses(endpoint: str, records: list[dict[str, Any]])
     seen: dict[str, dict[str, Any]] = {}
     for rec in records:
         body = rec["response"]["body"]
-        if endpoint == "/v1/models":
+        if endpoint in ("/v1/models", "/v1/openai/v1/models"):
             for m in body:
                 key = m.id
                 seen[key] = m
@@ -665,7 +666,7 @@ async def _patched_inference_method(original_method, self, client_type, endpoint
         logger.info(f"  Test context: {get_test_context()}")
 
     if mode == APIRecordingMode.LIVE or storage is None:
-        if endpoint == "/v1/models":
+        if endpoint in ("/v1/models", "/v1/openai/v1/models"):
             return original_method(self, *args, **kwargs)
         else:
             return await original_method(self, *args, **kwargs)
@@ -699,7 +700,7 @@ async def _patched_inference_method(original_method, self, client_type, endpoint
     recording = None
     if mode == APIRecordingMode.REPLAY or mode == APIRecordingMode.RECORD_IF_MISSING:
         # Special handling for model-list endpoints: merge all recordings with this hash
-        if endpoint in ("/api/tags", "/v1/models"):
+        if endpoint in ("/api/tags", "/v1/models", "/v1/openai/v1/models"):
             records = storage._model_list_responses(request_hash)
             recording = _combine_model_list_responses(endpoint, records)
         else:
@@ -739,13 +740,13 @@ async def _patched_inference_method(original_method, self, client_type, endpoint
             )
 
     if mode == APIRecordingMode.RECORD or (mode == APIRecordingMode.RECORD_IF_MISSING and not recording):
-        if endpoint == "/v1/models":
+        if endpoint in ("/v1/models", "/v1/openai/v1/models"):
             response = original_method(self, *args, **kwargs)
         else:
             response = await original_method(self, *args, **kwargs)
 
         # we want to store the result of the iterator, not the iterator itself
-        if endpoint == "/v1/models":
+        if endpoint in ("/v1/models", "/v1/openai/v1/models"):
             response = [m async for m in response]
 
         request_data = {

src/llama_stack/ui/app/api/v1/[...path]/route.ts

@@ -51,10 +51,14 @@ async function proxyRequest(request: NextRequest, method: string) {
     );
 
     // Create response with same status and headers
-    const proxyResponse = new NextResponse(responseText, {
-      status: response.status,
-      statusText: response.statusText,
-    });
+    // Handle 204 No Content responses specially
+    const proxyResponse =
+      response.status === 204
+        ? new NextResponse(null, { status: 204 })
+        : new NextResponse(responseText, {
+            status: response.status,
+            statusText: response.statusText,
+          });
 
     // Copy response headers (except problematic ones)
     response.headers.forEach((value, key) => {

src/llama_stack/ui/app/prompts/page.tsx

@@ -0,0 +1,5 @@
+import { PromptManagement } from "@/components/prompts";
+
+export default function PromptsPage() {
+  return <PromptManagement />;
+}

src/llama_stack/ui/components/layout/app-sidebar.tsx

@@ -8,6 +8,7 @@ import {
   MessageCircle,
   Settings2,
   Compass,
+  FileText,
 } from "lucide-react";
 import Link from "next/link";
 import { usePathname } from "next/navigation";
@@ -50,6 +51,11 @@ const manageItems = [
     url: "/logs/vector-stores",
     icon: Database,
   },
+  {
+    title: "Prompts",
+    url: "/prompts",
+    icon: FileText,
+  },
   {
     title: "Documentation",
     url: "https://llama-stack.readthedocs.io/en/latest/references/api_reference/index.html",

src/llama_stack/ui/components/prompts/index.ts

@@ -0,0 +1,4 @@
+export { PromptManagement } from "./prompt-management";
+export { PromptList } from "./prompt-list";
+export { PromptEditor } from "./prompt-editor";
+export * from "./types";

src/llama_stack/ui/components/prompts/prompt-editor.test.tsx

@@ -0,0 +1,309 @@
+import React from "react";
+import { render, screen, fireEvent } from "@testing-library/react";
+import "@testing-library/jest-dom";
+import { PromptEditor } from "./prompt-editor";
+import type { Prompt, PromptFormData } from "./types";
+
+describe("PromptEditor", () => {
+  const mockOnSave = jest.fn();
+  const mockOnCancel = jest.fn();
+  const mockOnDelete = jest.fn();
+
+  const defaultProps = {
+    onSave: mockOnSave,
+    onCancel: mockOnCancel,
+    onDelete: mockOnDelete,
+  };
+
+  beforeEach(() => {
+    jest.clearAllMocks();
+  });
+
+  describe("Create Mode", () => {
+    test("renders create form correctly", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      expect(screen.getByLabelText("Prompt Content *")).toBeInTheDocument();
+      expect(screen.getByText("Variables")).toBeInTheDocument();
+      expect(screen.getByText("Preview")).toBeInTheDocument();
+      expect(screen.getByText("Create Prompt")).toBeInTheDocument();
+      expect(screen.getByText("Cancel")).toBeInTheDocument();
+    });
+
+    test("shows preview placeholder when no content", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      expect(
+        screen.getByText("Enter content to preview the compiled prompt")
+      ).toBeInTheDocument();
+    });
+
+    test("submits form with correct data", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, {
+        target: { value: "Hello {{name}}, welcome!" },
+      });
+
+      fireEvent.click(screen.getByText("Create Prompt"));
+
+      expect(mockOnSave).toHaveBeenCalledWith({
+        prompt: "Hello {{name}}, welcome!",
+        variables: [],
+      });
+    });
+
+    test("prevents submission with empty prompt", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      fireEvent.click(screen.getByText("Create Prompt"));
+
+      expect(mockOnSave).not.toHaveBeenCalled();
+    });
+  });
+
+  describe("Edit Mode", () => {
+    const mockPrompt: Prompt = {
+      prompt_id: "prompt_123",
+      prompt: "Hello {{name}}, how is {{weather}}?",
+      version: 1,
+      variables: ["name", "weather"],
+      is_default: true,
+    };
+
+    test("renders edit form with existing data", () => {
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      expect(
+        screen.getByDisplayValue("Hello {{name}}, how is {{weather}}?")
+      ).toBeInTheDocument();
+      expect(screen.getAllByText("name")).toHaveLength(2); // One in variables, one in preview
+      expect(screen.getAllByText("weather")).toHaveLength(2); // One in variables, one in preview
+      expect(screen.getByText("Update Prompt")).toBeInTheDocument();
+      expect(screen.getByText("Delete Prompt")).toBeInTheDocument();
+    });
+
+    test("submits updated data correctly", () => {
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, {
+        target: { value: "Updated: Hello {{name}}!" },
+      });
+
+      fireEvent.click(screen.getByText("Update Prompt"));
+
+      expect(mockOnSave).toHaveBeenCalledWith({
+        prompt: "Updated: Hello {{name}}!",
+        variables: ["name", "weather"],
+      });
+    });
+  });
+
+  describe("Variables Management", () => {
+    test("adds new variable", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      const variableInput = screen.getByPlaceholderText(
+        "Add variable name (e.g. user_name, topic)"
+      );
+      fireEvent.change(variableInput, { target: { value: "testVar" } });
+      fireEvent.click(screen.getByText("Add"));
+
+      expect(screen.getByText("testVar")).toBeInTheDocument();
+    });
+
+    test("prevents adding duplicate variables", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      const variableInput = screen.getByPlaceholderText(
+        "Add variable name (e.g. user_name, topic)"
+      );
+
+      // Add first variable
+      fireEvent.change(variableInput, { target: { value: "test" } });
+      fireEvent.click(screen.getByText("Add"));
+
+      // Try to add same variable again
+      fireEvent.change(variableInput, { target: { value: "test" } });
+
+      // Button should be disabled
+      expect(screen.getByText("Add")).toBeDisabled();
+    });
+
+    test("removes variable", () => {
+      const mockPrompt: Prompt = {
+        prompt_id: "prompt_123",
+        prompt: "Hello {{name}}",
+        version: 1,
+        variables: ["name", "location"],
+        is_default: true,
+      };
+
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      // Check that both variables are present initially
+      expect(screen.getAllByText("name").length).toBeGreaterThan(0);
+      expect(screen.getAllByText("location").length).toBeGreaterThan(0);
+
+      // Remove the location variable by clicking the X button with the specific title
+      const removeLocationButton = screen.getByTitle(
+        "Remove location variable"
+      );
+      fireEvent.click(removeLocationButton);
+
+      // Name should still be there, location should be gone from the variables section
+      expect(screen.getAllByText("name").length).toBeGreaterThan(0);
+      expect(
+        screen.queryByTitle("Remove location variable")
+      ).not.toBeInTheDocument();
+    });
+
+    test("adds variable on Enter key", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      const variableInput = screen.getByPlaceholderText(
+        "Add variable name (e.g. user_name, topic)"
+      );
+      fireEvent.change(variableInput, { target: { value: "enterVar" } });
+
+      // Simulate Enter key press
+      fireEvent.keyPress(variableInput, {
+        key: "Enter",
+        code: "Enter",
+        charCode: 13,
+        preventDefault: jest.fn(),
+      });
+
+      // Check if the variable was added by looking for the badge
+      expect(screen.getAllByText("enterVar").length).toBeGreaterThan(0);
+    });
+  });
+
+  describe("Preview Functionality", () => {
+    test("shows live preview with variables", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      // Add prompt content
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, {
+        target: { value: "Hello {{name}}, welcome to {{place}}!" },
+      });
+
+      // Add variables
+      const variableInput = screen.getByPlaceholderText(
+        "Add variable name (e.g. user_name, topic)"
+      );
+      fireEvent.change(variableInput, { target: { value: "name" } });
+      fireEvent.click(screen.getByText("Add"));
+
+      fireEvent.change(variableInput, { target: { value: "place" } });
+      fireEvent.click(screen.getByText("Add"));
+
+      // Check that preview area shows the content
+      expect(screen.getByText("Compiled Prompt")).toBeInTheDocument();
+    });
+
+    test("shows variable value inputs in preview", () => {
+      const mockPrompt: Prompt = {
+        prompt_id: "prompt_123",
+        prompt: "Hello {{name}}",
+        version: 1,
+        variables: ["name"],
+        is_default: true,
+      };
+
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      expect(screen.getByText("Variable Values")).toBeInTheDocument();
+      expect(
+        screen.getByPlaceholderText("Enter value for name")
+      ).toBeInTheDocument();
+    });
+
+    test("shows color legend for variable states", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      // Add content to show preview
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, {
+        target: { value: "Hello {{name}}" },
+      });
+
+      expect(screen.getByText("Used")).toBeInTheDocument();
+      expect(screen.getByText("Unused")).toBeInTheDocument();
+      expect(screen.getByText("Undefined")).toBeInTheDocument();
+    });
+  });
+
+  describe("Error Handling", () => {
+    test("displays error message", () => {
+      const errorMessage = "Prompt contains undeclared variables";
+      render(<PromptEditor {...defaultProps} error={errorMessage} />);
+
+      expect(screen.getByText(errorMessage)).toBeInTheDocument();
+    });
+  });
+
+  describe("Delete Functionality", () => {
+    const mockPrompt: Prompt = {
+      prompt_id: "prompt_123",
+      prompt: "Hello {{name}}",
+      version: 1,
+      variables: ["name"],
+      is_default: true,
+    };
+
+    test("shows delete button in edit mode", () => {
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      expect(screen.getByText("Delete Prompt")).toBeInTheDocument();
+    });
+
+    test("hides delete button in create mode", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      expect(screen.queryByText("Delete Prompt")).not.toBeInTheDocument();
+    });
+
+    test("calls onDelete with confirmation", () => {
+      const originalConfirm = window.confirm;
+      window.confirm = jest.fn(() => true);
+
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      fireEvent.click(screen.getByText("Delete Prompt"));
+
+      expect(window.confirm).toHaveBeenCalledWith(
+        "Are you sure you want to delete this prompt? This action cannot be undone."
+      );
+      expect(mockOnDelete).toHaveBeenCalledWith("prompt_123");
+
+      window.confirm = originalConfirm;
+    });
+
+    test("does not delete when confirmation is cancelled", () => {
+      const originalConfirm = window.confirm;
+      window.confirm = jest.fn(() => false);
+
+      render(<PromptEditor {...defaultProps} prompt={mockPrompt} />);
+
+      fireEvent.click(screen.getByText("Delete Prompt"));
+
+      expect(mockOnDelete).not.toHaveBeenCalled();
+
+      window.confirm = originalConfirm;
+    });
+  });
+
+  describe("Cancel Functionality", () => {
+    test("calls onCancel when cancel button is clicked", () => {
+      render(<PromptEditor {...defaultProps} />);
+
+      fireEvent.click(screen.getByText("Cancel"));
+
+      expect(mockOnCancel).toHaveBeenCalled();
+    });
+  });
+});

src/llama_stack/ui/components/prompts/prompt-editor.tsx

@@ -0,0 +1,346 @@
+"use client";
+
+import { useState, useEffect } from "react";
+import { Button } from "@/components/ui/button";
+import { Input } from "@/components/ui/input";
+import { Label } from "@/components/ui/label";
+import { Textarea } from "@/components/ui/textarea";
+import { Badge } from "@/components/ui/badge";
+import {
+  Card,
+  CardContent,
+  CardDescription,
+  CardHeader,
+  CardTitle,
+} from "@/components/ui/card";
+import { Separator } from "@/components/ui/separator";
+import { X, Plus, Save, Trash2 } from "lucide-react";
+import { Prompt, PromptFormData } from "./types";
+
+interface PromptEditorProps {
+  prompt?: Prompt;
+  onSave: (prompt: PromptFormData) => void;
+  onCancel: () => void;
+  onDelete?: (promptId: string) => void;
+  error?: string | null;
+}
+
+export function PromptEditor({
+  prompt,
+  onSave,
+  onCancel,
+  onDelete,
+  error,
+}: PromptEditorProps) {
+  const [formData, setFormData] = useState<PromptFormData>({
+    prompt: "",
+    variables: [],
+  });
+
+  const [newVariable, setNewVariable] = useState("");
+  const [variableValues, setVariableValues] = useState<Record<string, string>>(
+    {}
+  );
+
+  useEffect(() => {
+    if (prompt) {
+      setFormData({
+        prompt: prompt.prompt || "",
+        variables: prompt.variables || [],
+      });
+    }
+  }, [prompt]);
+
+  const handleSubmit = (e: React.FormEvent) => {
+    e.preventDefault();
+    if (!formData.prompt.trim()) {
+      return;
+    }
+    onSave(formData);
+  };
+
+  const addVariable = () => {
+    if (
+      newVariable.trim() &&
+      !formData.variables.includes(newVariable.trim())
+    ) {
+      setFormData(prev => ({
+        ...prev,
+        variables: [...prev.variables, newVariable.trim()],
+      }));
+      setNewVariable("");
+    }
+  };
+
+  const removeVariable = (variableToRemove: string) => {
+    setFormData(prev => ({
+      ...prev,
+      variables: prev.variables.filter(
+        variable => variable !== variableToRemove
+      ),
+    }));
+  };
+
+  const renderPreview = () => {
+    const text = formData.prompt;
+    if (!text) return text;
+
+    // Split text by variable patterns and process each part
+    const parts = text.split(/(\{\{\s*\w+\s*\}\})/g);
+
+    return parts.map((part, index) => {
+      const variableMatch = part.match(/\{\{\s*(\w+)\s*\}\}/);
+      if (variableMatch) {
+        const variableName = variableMatch[1];
+        const isDefined = formData.variables.includes(variableName);
+        const value = variableValues[variableName];
+
+        if (!isDefined) {
+          // Variable not in variables list - likely a typo/bug (RED)
+          return (
+            <span
+              key={index}
+              className="bg-red-100 text-red-800 dark:bg-red-900 dark:text-red-200 px-1 rounded font-medium"
+            >
+              {part}
+            </span>
+          );
+        } else if (value && value.trim()) {
+          // Variable defined and has value - show the value (GREEN)
+          return (
+            <span
+              key={index}
+              className="bg-green-100 text-green-800 dark:bg-green-900 dark:text-green-200 px-1 rounded font-medium"
+            >
+              {value}
+            </span>
+          );
+        } else {
+          // Variable defined but empty (YELLOW)
+          return (
+            <span
+              key={index}
+              className="bg-yellow-100 text-yellow-800 dark:bg-yellow-900 dark:text-yellow-200 px-1 rounded font-medium"
+            >
+              {part}
+            </span>
+          );
+        }
+      }
+      return part;
+    });
+  };
+
+  const updateVariableValue = (variable: string, value: string) => {
+    setVariableValues(prev => ({
+      ...prev,
+      [variable]: value,
+    }));
+  };
+
+  return (
+    <form onSubmit={handleSubmit} className="space-y-6">
+      {error && (
+        <div className="p-4 bg-destructive/10 border border-destructive/20 rounded-md">
+          <p className="text-destructive text-sm">{error}</p>
+        </div>
+      )}
+      <div className="grid grid-cols-1 lg:grid-cols-2 gap-6">
+        {/* Form Section */}
+        <div className="space-y-4">
+          <div>
+            <Label htmlFor="prompt">Prompt Content *</Label>
+            <Textarea
+              id="prompt"
+              value={formData.prompt}
+              onChange={e =>
+                setFormData(prev => ({ ...prev, prompt: e.target.value }))
+              }
+              placeholder="Enter your prompt content here. Use {{variable_name}} for dynamic variables."
+              className="min-h-32 font-mono mt-2"
+              required
+            />
+            <p className="text-xs text-muted-foreground mt-2">
+              Use double curly braces around variable names, e.g.,{" "}
+              {`{{user_name}}`} or {`{{topic}}`}
+            </p>
+          </div>
+
+          <div className="space-y-3">
+            <Label className="text-sm font-medium">Variables</Label>
+
+            <div className="flex gap-2 mt-2">
+              <Input
+                value={newVariable}
+                onChange={e => setNewVariable(e.target.value)}
+                placeholder="Add variable name (e.g. user_name, topic)"
+                onKeyPress={e =>
+                  e.key === "Enter" && (e.preventDefault(), addVariable())
+                }
+                className="flex-1"
+              />
+              <Button
+                type="button"
+                onClick={addVariable}
+                size="sm"
+                disabled={
+                  !newVariable.trim() ||
+                  formData.variables.includes(newVariable.trim())
+                }
+              >
+                <Plus className="h-4 w-4" />
+                Add
+              </Button>
+            </div>
+
+            {formData.variables.length > 0 && (
+              <div className="border rounded-lg p-3 bg-muted/20">
+                <div className="flex flex-wrap gap-2">
+                  {formData.variables.map(variable => (
+                    <Badge
+                      key={variable}
+                      variant="secondary"
+                      className="text-sm px-2 py-1"
+                    >
+                      {variable}
+                      <button
+                        type="button"
+                        onClick={() => removeVariable(variable)}
+                        className="ml-2 hover:text-destructive transition-colors"
+                        title={`Remove ${variable} variable`}
+                      >
+                        <X className="h-3 w-3" />
+                      </button>
+                    </Badge>
+                  ))}
+                </div>
+              </div>
+            )}
+
+            <p className="text-xs text-muted-foreground">
+              Variables that can be used in the prompt template. Each variable
+              should match a {`{{variable}}`} placeholder in the content above.
+            </p>
+          </div>
+        </div>
+
+        {/* Preview Section */}
+        <div className="space-y-4">
+          <Card>
+            <CardHeader>
+              <CardTitle className="text-lg">Preview</CardTitle>
+              <CardDescription>
+                Live preview of compiled prompt and variable substitution.
+              </CardDescription>
+            </CardHeader>
+            <CardContent className="space-y-4">
+              {formData.prompt ? (
+                <>
+                  {/* Variable Values */}
+                  {formData.variables.length > 0 && (
+                    <div className="space-y-3">
+                      <Label className="text-sm font-medium">
+                        Variable Values
+                      </Label>
+                      <div className="space-y-2">
+                        {formData.variables.map(variable => (
+                          <div
+                            key={variable}
+                            className="grid grid-cols-2 gap-3 items-center"
+                          >
+                            <div className="text-sm font-mono text-muted-foreground">
+                              {variable}
+                            </div>
+                            <Input
+                              id={`var-${variable}`}
+                              value={variableValues[variable] || ""}
+                              onChange={e =>
+                                updateVariableValue(variable, e.target.value)
+                              }
+                              placeholder={`Enter value for ${variable}`}
+                              className="text-sm"
+                            />
+                          </div>
+                        ))}
+                      </div>
+                      <Separator />
+                    </div>
+                  )}
+
+                  {/* Live Preview */}
+                  <div>
+                    <Label className="text-sm font-medium mb-2 block">
+                      Compiled Prompt
+                    </Label>
+                    <div className="bg-muted/50 p-4 rounded-lg border">
+                      <div className="text-sm leading-relaxed whitespace-pre-wrap">
+                        {renderPreview()}
+                      </div>
+                    </div>
+                    <div className="flex flex-wrap gap-4 mt-2 text-xs">
+                      <div className="flex items-center gap-1">
+                        <div className="w-3 h-3 bg-green-500 dark:bg-green-400 border rounded"></div>
+                        <span className="text-muted-foreground">Used</span>
+                      </div>
+                      <div className="flex items-center gap-1">
+                        <div className="w-3 h-3 bg-yellow-500 dark:bg-yellow-400 border rounded"></div>
+                        <span className="text-muted-foreground">Unused</span>
+                      </div>
+                      <div className="flex items-center gap-1">
+                        <div className="w-3 h-3 bg-red-500 dark:bg-red-400 border rounded"></div>
+                        <span className="text-muted-foreground">Undefined</span>
+                      </div>
+                    </div>
+                  </div>
+                </>
+              ) : (
+                <div className="text-center py-8">
+                  <div className="text-muted-foreground text-sm">
+                    Enter content to preview the compiled prompt
+                  </div>
+                  <div className="text-xs text-muted-foreground mt-2">
+                    Use {`{{variable_name}}`} to add dynamic variables
+                  </div>
+                </div>
+              )}
+            </CardContent>
+          </Card>
+        </div>
+      </div>
+
+      <Separator />
+
+      <div className="flex justify-between">
+        <div>
+          {prompt && onDelete && (
+            <Button
+              type="button"
+              variant="destructive"
+              onClick={() => {
+                if (
+                  confirm(
+                    `Are you sure you want to delete this prompt? This action cannot be undone.`
+                  )
+                ) {
+                  onDelete(prompt.prompt_id);
+                }
+              }}
+            >
+              <Trash2 className="h-4 w-4 mr-2" />
+              Delete Prompt
+            </Button>
+          )}
+        </div>
+        <div className="flex gap-2">
+          <Button type="button" variant="outline" onClick={onCancel}>
+            Cancel
+          </Button>
+          <Button type="submit">
+            <Save className="h-4 w-4 mr-2" />
+            {prompt ? "Update" : "Create"} Prompt
+          </Button>
+        </div>
+      </div>
+    </form>
+  );
+}

src/llama_stack/ui/components/prompts/prompt-list.test.tsx

@@ -0,0 +1,259 @@
+import React from "react";
+import { render, screen, fireEvent } from "@testing-library/react";
+import "@testing-library/jest-dom";
+import { PromptList } from "./prompt-list";
+import type { Prompt } from "./types";
+
+describe("PromptList", () => {
+  const mockOnEdit = jest.fn();
+  const mockOnDelete = jest.fn();
+
+  const defaultProps = {
+    prompts: [],
+    onEdit: mockOnEdit,
+    onDelete: mockOnDelete,
+  };
+
+  beforeEach(() => {
+    jest.clearAllMocks();
+  });
+
+  describe("Empty State", () => {
+    test("renders empty message when no prompts", () => {
+      render(<PromptList {...defaultProps} />);
+
+      expect(screen.getByText("No prompts yet")).toBeInTheDocument();
+    });
+
+    test("shows filtered empty message when search has no results", () => {
+      const prompts: Prompt[] = [
+        {
+          prompt_id: "prompt_123",
+          prompt: "Hello world",
+          version: 1,
+          variables: [],
+          is_default: false,
+        },
+      ];
+
+      render(<PromptList {...defaultProps} prompts={prompts} />);
+
+      // Search for something that doesn't exist
+      const searchInput = screen.getByPlaceholderText("Search prompts...");
+      fireEvent.change(searchInput, { target: { value: "nonexistent" } });
+
+      expect(
+        screen.getByText("No prompts match your filters")
+      ).toBeInTheDocument();
+    });
+  });
+
+  describe("Prompts Display", () => {
+    const mockPrompts: Prompt[] = [
+      {
+        prompt_id: "prompt_123",
+        prompt: "Hello {{name}}, how are you?",
+        version: 1,
+        variables: ["name"],
+        is_default: true,
+      },
+      {
+        prompt_id: "prompt_456",
+        prompt: "Summarize this {{text}} in {{length}} words",
+        version: 2,
+        variables: ["text", "length"],
+        is_default: false,
+      },
+      {
+        prompt_id: "prompt_789",
+        prompt: "Simple prompt with no variables",
+        version: 1,
+        variables: [],
+        is_default: false,
+      },
+    ];
+
+    test("renders prompts table with correct headers", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      expect(screen.getByText("ID")).toBeInTheDocument();
+      expect(screen.getByText("Content")).toBeInTheDocument();
+      expect(screen.getByText("Variables")).toBeInTheDocument();
+      expect(screen.getByText("Version")).toBeInTheDocument();
+      expect(screen.getByText("Actions")).toBeInTheDocument();
+    });
+
+    test("renders prompt data correctly", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      // Check prompt IDs
+      expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      expect(screen.getByText("prompt_456")).toBeInTheDocument();
+      expect(screen.getByText("prompt_789")).toBeInTheDocument();
+
+      // Check content
+      expect(
+        screen.getByText("Hello {{name}}, how are you?")
+      ).toBeInTheDocument();
+      expect(
+        screen.getByText("Summarize this {{text}} in {{length}} words")
+      ).toBeInTheDocument();
+      expect(
+        screen.getByText("Simple prompt with no variables")
+      ).toBeInTheDocument();
+
+      // Check versions
+      expect(screen.getAllByText("1")).toHaveLength(2); // Two prompts with version 1
+      expect(screen.getByText("2")).toBeInTheDocument();
+
+      // Check default badge
+      expect(screen.getByText("Default")).toBeInTheDocument();
+    });
+
+    test("renders variables correctly", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      // Check variables display
+      expect(screen.getByText("name")).toBeInTheDocument();
+      expect(screen.getByText("text")).toBeInTheDocument();
+      expect(screen.getByText("length")).toBeInTheDocument();
+      expect(screen.getByText("None")).toBeInTheDocument(); // For prompt with no variables
+    });
+
+    test("prompt ID links are clickable and call onEdit", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      // Click on the first prompt ID link
+      const promptLink = screen.getByRole("button", { name: "prompt_123" });
+      fireEvent.click(promptLink);
+
+      expect(mockOnEdit).toHaveBeenCalledWith(mockPrompts[0]);
+    });
+
+    test("edit buttons call onEdit", () => {
+      const { container } = render(
+        <PromptList {...defaultProps} prompts={mockPrompts} />
+      );
+
+      // Find the action buttons in the table - they should be in the last column
+      const actionCells = container.querySelectorAll("td:last-child");
+      const firstActionCell = actionCells[0];
+      const editButton = firstActionCell?.querySelector("button");
+
+      expect(editButton).toBeInTheDocument();
+      fireEvent.click(editButton!);
+
+      expect(mockOnEdit).toHaveBeenCalledWith(mockPrompts[0]);
+    });
+
+    test("delete buttons call onDelete with confirmation", () => {
+      const originalConfirm = window.confirm;
+      window.confirm = jest.fn(() => true);
+
+      const { container } = render(
+        <PromptList {...defaultProps} prompts={mockPrompts} />
+      );
+
+      // Find the delete button (second button in the first action cell)
+      const actionCells = container.querySelectorAll("td:last-child");
+      const firstActionCell = actionCells[0];
+      const buttons = firstActionCell?.querySelectorAll("button");
+      const deleteButton = buttons?.[1]; // Second button should be delete
+
+      expect(deleteButton).toBeInTheDocument();
+      fireEvent.click(deleteButton!);
+
+      expect(window.confirm).toHaveBeenCalledWith(
+        "Are you sure you want to delete this prompt? This action cannot be undone."
+      );
+      expect(mockOnDelete).toHaveBeenCalledWith("prompt_123");
+
+      window.confirm = originalConfirm;
+    });
+
+    test("delete does not execute when confirmation is cancelled", () => {
+      const originalConfirm = window.confirm;
+      window.confirm = jest.fn(() => false);
+
+      const { container } = render(
+        <PromptList {...defaultProps} prompts={mockPrompts} />
+      );
+
+      const actionCells = container.querySelectorAll("td:last-child");
+      const firstActionCell = actionCells[0];
+      const buttons = firstActionCell?.querySelectorAll("button");
+      const deleteButton = buttons?.[1]; // Second button should be delete
+
+      expect(deleteButton).toBeInTheDocument();
+      fireEvent.click(deleteButton!);
+
+      expect(mockOnDelete).not.toHaveBeenCalled();
+
+      window.confirm = originalConfirm;
+    });
+  });
+
+  describe("Search Functionality", () => {
+    const mockPrompts: Prompt[] = [
+      {
+        prompt_id: "user_greeting",
+        prompt: "Hello {{name}}, welcome!",
+        version: 1,
+        variables: ["name"],
+        is_default: true,
+      },
+      {
+        prompt_id: "system_summary",
+        prompt: "Summarize the following text",
+        version: 1,
+        variables: [],
+        is_default: false,
+      },
+    ];
+
+    test("filters prompts by prompt ID", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      const searchInput = screen.getByPlaceholderText("Search prompts...");
+      fireEvent.change(searchInput, { target: { value: "user" } });
+
+      expect(screen.getByText("user_greeting")).toBeInTheDocument();
+      expect(screen.queryByText("system_summary")).not.toBeInTheDocument();
+    });
+
+    test("filters prompts by content", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      const searchInput = screen.getByPlaceholderText("Search prompts...");
+      fireEvent.change(searchInput, { target: { value: "welcome" } });
+
+      expect(screen.getByText("user_greeting")).toBeInTheDocument();
+      expect(screen.queryByText("system_summary")).not.toBeInTheDocument();
+    });
+
+    test("search is case insensitive", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      const searchInput = screen.getByPlaceholderText("Search prompts...");
+      fireEvent.change(searchInput, { target: { value: "HELLO" } });
+
+      expect(screen.getByText("user_greeting")).toBeInTheDocument();
+      expect(screen.queryByText("system_summary")).not.toBeInTheDocument();
+    });
+
+    test("clearing search shows all prompts", () => {
+      render(<PromptList {...defaultProps} prompts={mockPrompts} />);
+
+      const searchInput = screen.getByPlaceholderText("Search prompts...");
+
+      // Filter first
+      fireEvent.change(searchInput, { target: { value: "user" } });
+      expect(screen.queryByText("system_summary")).not.toBeInTheDocument();
+
+      // Clear search
+      fireEvent.change(searchInput, { target: { value: "" } });
+      expect(screen.getByText("user_greeting")).toBeInTheDocument();
+      expect(screen.getByText("system_summary")).toBeInTheDocument();
+    });
+  });
+});

src/llama_stack/ui/components/prompts/prompt-list.tsx

@@ -0,0 +1,164 @@
+"use client";
+
+import { useState } from "react";
+import { Badge } from "@/components/ui/badge";
+import { Button } from "@/components/ui/button";
+import {
+  Table,
+  TableBody,
+  TableCell,
+  TableHead,
+  TableHeader,
+  TableRow,
+} from "@/components/ui/table";
+import { Input } from "@/components/ui/input";
+import { Edit, Search, Trash2 } from "lucide-react";
+import { Prompt, PromptFilters } from "./types";
+
+interface PromptListProps {
+  prompts: Prompt[];
+  onEdit: (prompt: Prompt) => void;
+  onDelete: (promptId: string) => void;
+}
+
+export function PromptList({ prompts, onEdit, onDelete }: PromptListProps) {
+  const [filters, setFilters] = useState<PromptFilters>({});
+
+  const filteredPrompts = prompts.filter(prompt => {
+    if (
+      filters.searchTerm &&
+      !(
+        prompt.prompt
+          ?.toLowerCase()
+          .includes(filters.searchTerm.toLowerCase()) ||
+        prompt.prompt_id
+          .toLowerCase()
+          .includes(filters.searchTerm.toLowerCase())
+      )
+    ) {
+      return false;
+    }
+    return true;
+  });
+
+  return (
+    <div className="space-y-4">
+      {/* Filters */}
+      <div className="flex flex-col sm:flex-row gap-4">
+        <div className="relative flex-1">
+          <Search className="absolute left-3 top-1/2 transform -translate-y-1/2 text-muted-foreground h-4 w-4" />
+          <Input
+            placeholder="Search prompts..."
+            value={filters.searchTerm || ""}
+            onChange={e =>
+              setFilters(prev => ({ ...prev, searchTerm: e.target.value }))
+            }
+            className="pl-10"
+          />
+        </div>
+      </div>
+
+      {/* Prompts Table */}
+      <div className="overflow-auto">
+        <Table>
+          <TableHeader>
+            <TableRow>
+              <TableHead>ID</TableHead>
+              <TableHead>Content</TableHead>
+              <TableHead>Variables</TableHead>
+              <TableHead>Version</TableHead>
+              <TableHead>Actions</TableHead>
+            </TableRow>
+          </TableHeader>
+          <TableBody>
+            {filteredPrompts.map(prompt => (
+              <TableRow key={prompt.prompt_id}>
+                <TableCell className="max-w-48">
+                  <Button
+                    variant="link"
+                    className="p-0 h-auto font-mono text-blue-600 hover:text-blue-800 dark:text-blue-400 dark:hover:text-blue-300 max-w-full justify-start"
+                    onClick={() => onEdit(prompt)}
+                    title={prompt.prompt_id}
+                  >
+                    <div className="truncate">{prompt.prompt_id}</div>
+                  </Button>
+                </TableCell>
+                <TableCell className="max-w-64">
+                  <div
+                    className="font-mono text-xs text-muted-foreground truncate"
+                    title={prompt.prompt || "No content"}
+                  >
+                    {prompt.prompt || "No content"}
+                  </div>
+                </TableCell>
+                <TableCell>
+                  {prompt.variables.length > 0 ? (
+                    <div className="flex flex-wrap gap-1">
+                      {prompt.variables.map(variable => (
+                        <Badge
+                          key={variable}
+                          variant="outline"
+                          className="text-xs"
+                        >
+                          {variable}
+                        </Badge>
+                      ))}
+                    </div>
+                  ) : (
+                    <span className="text-muted-foreground text-sm">None</span>
+                  )}
+                </TableCell>
+                <TableCell className="text-sm">
+                  {prompt.version}
+                  {prompt.is_default && (
+                    <Badge variant="secondary" className="text-xs ml-2">
+                      Default
+                    </Badge>
+                  )}
+                </TableCell>
+                <TableCell>
+                  <div className="flex gap-1">
+                    <Button
+                      size="sm"
+                      variant="outline"
+                      onClick={() => onEdit(prompt)}
+                      className="h-8 w-8 p-0"
+                    >
+                      <Edit className="h-3 w-3" />
+                    </Button>
+                    <Button
+                      size="sm"
+                      variant="outline"
+                      onClick={() => {
+                        if (
+                          confirm(
+                            `Are you sure you want to delete this prompt? This action cannot be undone.`
+                          )
+                        ) {
+                          onDelete(prompt.prompt_id);
+                        }
+                      }}
+                      className="h-8 w-8 p-0 text-destructive hover:text-destructive"
+                    >
+                      <Trash2 className="h-3 w-3" />
+                    </Button>
+                  </div>
+                </TableCell>
+              </TableRow>
+            ))}
+          </TableBody>
+        </Table>
+      </div>
+
+      {filteredPrompts.length === 0 && (
+        <div className="text-center py-12">
+          <div className="text-muted-foreground">
+            {prompts.length === 0
+              ? "No prompts yet"
+              : "No prompts match your filters"}
+          </div>
+        </div>
+      )}
+    </div>
+  );
+}

src/llama_stack/ui/components/prompts/prompt-management.test.tsx

@@ -0,0 +1,304 @@
+import React from "react";
+import { render, screen, fireEvent, waitFor } from "@testing-library/react";
+import "@testing-library/jest-dom";
+import { PromptManagement } from "./prompt-management";
+import type { Prompt } from "./types";
+
+// Mock the auth client
+const mockPromptsClient = {
+  list: jest.fn(),
+  create: jest.fn(),
+  update: jest.fn(),
+  delete: jest.fn(),
+};
+
+jest.mock("@/hooks/use-auth-client", () => ({
+  useAuthClient: () => ({
+    prompts: mockPromptsClient,
+  }),
+}));
+
+describe("PromptManagement", () => {
+  beforeEach(() => {
+    jest.clearAllMocks();
+  });
+
+  describe("Loading State", () => {
+    test("renders loading state initially", () => {
+      mockPromptsClient.list.mockReturnValue(new Promise(() => {})); // Never resolves
+      render(<PromptManagement />);
+
+      expect(screen.getByText("Loading prompts...")).toBeInTheDocument();
+      expect(screen.getByText("Prompts")).toBeInTheDocument();
+    });
+  });
+
+  describe("Empty State", () => {
+    test("renders empty state when no prompts", async () => {
+      mockPromptsClient.list.mockResolvedValue([]);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("No prompts found.")).toBeInTheDocument();
+      });
+
+      expect(screen.getByText("Create Your First Prompt")).toBeInTheDocument();
+    });
+
+    test("opens modal when clicking 'Create Your First Prompt'", async () => {
+      mockPromptsClient.list.mockResolvedValue([]);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(
+          screen.getByText("Create Your First Prompt")
+        ).toBeInTheDocument();
+      });
+
+      fireEvent.click(screen.getByText("Create Your First Prompt"));
+
+      expect(screen.getByText("Create New Prompt")).toBeInTheDocument();
+    });
+  });
+
+  describe("Error State", () => {
+    test("renders error state when API fails", async () => {
+      const error = new Error("API not found");
+      mockPromptsClient.list.mockRejectedValue(error);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText(/Error:/)).toBeInTheDocument();
+      });
+    });
+
+    test("renders specific error for 404", async () => {
+      const error = new Error("404 Not found");
+      mockPromptsClient.list.mockRejectedValue(error);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(
+          screen.getByText(/Prompts API endpoint not found/)
+        ).toBeInTheDocument();
+      });
+    });
+  });
+
+  describe("Prompts List", () => {
+    const mockPrompts: Prompt[] = [
+      {
+        prompt_id: "prompt_123",
+        prompt: "Hello {{name}}, how are you?",
+        version: 1,
+        variables: ["name"],
+        is_default: true,
+      },
+      {
+        prompt_id: "prompt_456",
+        prompt: "Summarize this {{text}}",
+        version: 2,
+        variables: ["text"],
+        is_default: false,
+      },
+    ];
+
+    test("renders prompts list correctly", async () => {
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      expect(screen.getByText("prompt_456")).toBeInTheDocument();
+      expect(
+        screen.getByText("Hello {{name}}, how are you?")
+      ).toBeInTheDocument();
+      expect(screen.getByText("Summarize this {{text}}")).toBeInTheDocument();
+    });
+
+    test("opens modal when clicking 'New Prompt' button", async () => {
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      fireEvent.click(screen.getByText("New Prompt"));
+
+      expect(screen.getByText("Create New Prompt")).toBeInTheDocument();
+    });
+  });
+
+  describe("Modal Operations", () => {
+    const mockPrompts: Prompt[] = [
+      {
+        prompt_id: "prompt_123",
+        prompt: "Hello {{name}}",
+        version: 1,
+        variables: ["name"],
+        is_default: true,
+      },
+    ];
+
+    test("closes modal when clicking cancel", async () => {
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      // Open modal
+      fireEvent.click(screen.getByText("New Prompt"));
+      expect(screen.getByText("Create New Prompt")).toBeInTheDocument();
+
+      // Close modal
+      fireEvent.click(screen.getByText("Cancel"));
+      expect(screen.queryByText("Create New Prompt")).not.toBeInTheDocument();
+    });
+
+    test("creates new prompt successfully", async () => {
+      const newPrompt: Prompt = {
+        prompt_id: "prompt_new",
+        prompt: "New prompt content",
+        version: 1,
+        variables: [],
+        is_default: false,
+      };
+
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      mockPromptsClient.create.mockResolvedValue(newPrompt);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      // Open modal
+      fireEvent.click(screen.getByText("New Prompt"));
+
+      // Fill form
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, {
+        target: { value: "New prompt content" },
+      });
+
+      // Submit form
+      fireEvent.click(screen.getByText("Create Prompt"));
+
+      await waitFor(() => {
+        expect(mockPromptsClient.create).toHaveBeenCalledWith({
+          prompt: "New prompt content",
+          variables: [],
+        });
+      });
+    });
+
+    test("handles create error gracefully", async () => {
+      const error = {
+        detail: {
+          errors: [{ msg: "Prompt contains undeclared variables: ['test']" }],
+        },
+      };
+
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      mockPromptsClient.create.mockRejectedValue(error);
+      render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      // Open modal
+      fireEvent.click(screen.getByText("New Prompt"));
+
+      // Fill form
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, { target: { value: "Hello {{test}}" } });
+
+      // Submit form
+      fireEvent.click(screen.getByText("Create Prompt"));
+
+      await waitFor(() => {
+        expect(
+          screen.getByText("Prompt contains undeclared variables: ['test']")
+        ).toBeInTheDocument();
+      });
+    });
+
+    test("updates existing prompt successfully", async () => {
+      const updatedPrompt: Prompt = {
+        ...mockPrompts[0],
+        prompt: "Updated content",
+      };
+
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      mockPromptsClient.update.mockResolvedValue(updatedPrompt);
+      const { container } = render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      // Click edit button (first button in the action cell of the first row)
+      const actionCells = container.querySelectorAll("td:last-child");
+      const firstActionCell = actionCells[0];
+      const editButton = firstActionCell?.querySelector("button");
+
+      expect(editButton).toBeInTheDocument();
+      fireEvent.click(editButton!);
+
+      expect(screen.getByText("Edit Prompt")).toBeInTheDocument();
+
+      // Update content
+      const promptInput = screen.getByLabelText("Prompt Content *");
+      fireEvent.change(promptInput, { target: { value: "Updated content" } });
+
+      // Submit form
+      fireEvent.click(screen.getByText("Update Prompt"));
+
+      await waitFor(() => {
+        expect(mockPromptsClient.update).toHaveBeenCalledWith("prompt_123", {
+          prompt: "Updated content",
+          variables: ["name"],
+          version: 1,
+          set_as_default: true,
+        });
+      });
+    });
+
+    test("deletes prompt successfully", async () => {
+      mockPromptsClient.list.mockResolvedValue(mockPrompts);
+      mockPromptsClient.delete.mockResolvedValue(undefined);
+
+      // Mock window.confirm
+      const originalConfirm = window.confirm;
+      window.confirm = jest.fn(() => true);
+
+      const { container } = render(<PromptManagement />);
+
+      await waitFor(() => {
+        expect(screen.getByText("prompt_123")).toBeInTheDocument();
+      });
+
+      // Click delete button (second button in the action cell of the first row)
+      const actionCells = container.querySelectorAll("td:last-child");
+      const firstActionCell = actionCells[0];
+      const buttons = firstActionCell?.querySelectorAll("button");
+      const deleteButton = buttons?.[1]; // Second button should be delete
+
+      expect(deleteButton).toBeInTheDocument();
+      fireEvent.click(deleteButton!);
+
+      await waitFor(() => {
+        expect(mockPromptsClient.delete).toHaveBeenCalledWith("prompt_123");
+      });
+
+      // Restore window.confirm
+      window.confirm = originalConfirm;
+    });
+  });
+});

src/llama_stack/ui/components/prompts/prompt-management.tsx

@@ -0,0 +1,233 @@
+"use client";
+
+import { useState, useEffect } from "react";
+import { Button } from "@/components/ui/button";
+import { Plus } from "lucide-react";
+import { PromptList } from "./prompt-list";
+import { PromptEditor } from "./prompt-editor";
+import { Prompt, PromptFormData } from "./types";
+import { useAuthClient } from "@/hooks/use-auth-client";
+
+export function PromptManagement() {
+  const [prompts, setPrompts] = useState<Prompt[]>([]);
+  const [showPromptModal, setShowPromptModal] = useState(false);
+  const [editingPrompt, setEditingPrompt] = useState<Prompt | undefined>();
+  const [loading, setLoading] = useState(true);
+  const [error, setError] = useState<string | null>(null); // For main page errors (loading, etc.)
+  const [modalError, setModalError] = useState<string | null>(null); // For form submission errors
+  const client = useAuthClient();
+
+  // Load prompts from API on component mount
+  useEffect(() => {
+    const fetchPrompts = async () => {
+      try {
+        setLoading(true);
+        setError(null);
+
+        const response = await client.prompts.list();
+        setPrompts(response || []);
+      } catch (err: unknown) {
+        console.error("Failed to load prompts:", err);
+
+        // Handle different types of errors
+        const error = err as Error & { status?: number };
+        if (error?.message?.includes("404") || error?.status === 404) {
+          setError(
+            "Prompts API endpoint not found. Please ensure your Llama Stack server supports the prompts API."
+          );
+        } else if (
+          error?.message?.includes("not implemented") ||
+          error?.message?.includes("not supported")
+        ) {
+          setError(
+            "Prompts API is not yet implemented on this Llama Stack server."
+          );
+        } else {
+          setError(
+            `Failed to load prompts: ${error?.message || "Unknown error"}`
+          );
+        }
+      } finally {
+        setLoading(false);
+      }
+    };
+
+    fetchPrompts();
+  }, [client]);
+
+  const handleSavePrompt = async (formData: PromptFormData) => {
+    try {
+      setModalError(null);
+
+      if (editingPrompt) {
+        // Update existing prompt
+        const response = await client.prompts.update(editingPrompt.prompt_id, {
+          prompt: formData.prompt,
+          variables: formData.variables,
+          version: editingPrompt.version,
+          set_as_default: true,
+        });
+
+        // Update local state
+        setPrompts(prev =>
+          prev.map(p =>
+            p.prompt_id === editingPrompt.prompt_id ? response : p
+          )
+        );
+      } else {
+        // Create new prompt
+        const response = await client.prompts.create({
+          prompt: formData.prompt,
+          variables: formData.variables,
+        });
+
+        // Add to local state
+        setPrompts(prev => [response, ...prev]);
+      }
+
+      setShowPromptModal(false);
+      setEditingPrompt(undefined);
+    } catch (err) {
+      console.error("Failed to save prompt:", err);
+
+      // Extract specific error message from API response
+      const error = err as Error & {
+        message?: string;
+        detail?: { errors?: Array<{ msg?: string }> };
+      };
+
+      // Try to parse JSON from error message if it's a string
+      let parsedError = error;
+      if (typeof error?.message === "string" && error.message.includes("{")) {
+        try {
+          const jsonMatch = error.message.match(/\d+\s+(.+)/);
+          if (jsonMatch) {
+            parsedError = JSON.parse(jsonMatch[1]);
+          }
+        } catch {
+          // If parsing fails, use original error
+        }
+      }
+
+      // Try to get the specific validation error message
+      const validationError = parsedError?.detail?.errors?.[0]?.msg;
+      if (validationError) {
+        // Clean up validation error messages (remove "Value error, " prefix if present)
+        const cleanMessage = validationError.replace(/^Value error,\s*/i, "");
+        setModalError(cleanMessage);
+      } else {
+        // For other errors, format them nicely with line breaks
+        const statusMatch = error?.message?.match(/(\d+)\s+(.+)/);
+        if (statusMatch) {
+          const statusCode = statusMatch[1];
+          const response = statusMatch[2];
+          setModalError(
+            `Failed to save prompt: Status Code ${statusCode}\n\nResponse: ${response}`
+          );
+        } else {
+          const message = error?.message || error?.detail || "Unknown error";
+          setModalError(`Failed to save prompt: ${message}`);
+        }
+      }
+    }
+  };
+
+  const handleEditPrompt = (prompt: Prompt) => {
+    setEditingPrompt(prompt);
+    setShowPromptModal(true);
+    setModalError(null); // Clear any previous modal errors
+  };
+
+  const handleDeletePrompt = async (promptId: string) => {
+    try {
+      setError(null);
+      await client.prompts.delete(promptId);
+      setPrompts(prev => prev.filter(p => p.prompt_id !== promptId));
+
+      // If we're deleting the currently editing prompt, close the modal
+      if (editingPrompt && editingPrompt.prompt_id === promptId) {
+        setShowPromptModal(false);
+        setEditingPrompt(undefined);
+      }
+    } catch (err) {
+      console.error("Failed to delete prompt:", err);
+      setError("Failed to delete prompt");
+    }
+  };
+
+  const handleCreateNew = () => {
+    setEditingPrompt(undefined);
+    setShowPromptModal(true);
+    setModalError(null); // Clear any previous modal errors
+  };
+
+  const handleCancel = () => {
+    setShowPromptModal(false);
+    setEditingPrompt(undefined);
+  };
+
+  const renderContent = () => {
+    if (loading) {
+      return <div className="text-muted-foreground">Loading prompts...</div>;
+    }
+
+    if (error) {
+      return <div className="text-destructive">Error: {error}</div>;
+    }
+
+    if (!prompts || prompts.length === 0) {
+      return (
+        <div className="text-center py-12">
+          <p className="text-muted-foreground mb-4">No prompts found.</p>
+          <Button onClick={handleCreateNew}>
+            <Plus className="h-4 w-4 mr-2" />
+            Create Your First Prompt
+          </Button>
+        </div>
+      );
+    }
+
+    return (
+      <PromptList
+        prompts={prompts}
+        onEdit={handleEditPrompt}
+        onDelete={handleDeletePrompt}
+      />
+    );
+  };
+
+  return (
+    <div className="space-y-4">
+      <div className="flex items-center justify-between">
+        <h1 className="text-2xl font-semibold">Prompts</h1>
+        <Button onClick={handleCreateNew} disabled={loading}>
+          <Plus className="h-4 w-4 mr-2" />
+          New Prompt
+        </Button>
+      </div>
+      {renderContent()}
+
+      {/* Create/Edit Prompt Modal */}
+      {showPromptModal && (
+        <div className="fixed inset-0 bg-black/50 flex items-center justify-center z-50">
+          <div className="bg-background border rounded-lg shadow-lg max-w-4xl w-full mx-4 max-h-[90vh] overflow-hidden">
+            <div className="p-6 border-b">
+              <h2 className="text-2xl font-bold">
+                {editingPrompt ? "Edit Prompt" : "Create New Prompt"}
+              </h2>
+            </div>
+            <div className="p-6 overflow-y-auto max-h-[calc(90vh-120px)]">
+              <PromptEditor
+                prompt={editingPrompt}
+                onSave={handleSavePrompt}
+                onCancel={handleCancel}
+                onDelete={handleDeletePrompt}
+                error={modalError}
+              />
+            </div>
+          </div>
+        </div>
+      )}
+    </div>
+  );
+}

src/llama_stack/ui/components/prompts/types.ts

@@ -0,0 +1,16 @@
+export interface Prompt {
+  prompt_id: string;
+  prompt: string | null;
+  version: number;
+  variables: string[];
+  is_default: boolean;
+}
+
+export interface PromptFormData {
+  prompt: string;
+  variables: string[];
+}
+
+export interface PromptFilters {
+  searchTerm?: string;
+}

src/llama_stack/ui/components/ui/badge.tsx

@@ -0,0 +1,36 @@
+import * as React from "react";
+import { cva, type VariantProps } from "class-variance-authority";
+
+import { cn } from "@/lib/utils";
+
+const badgeVariants = cva(
+  "inline-flex items-center rounded-full border px-2.5 py-0.5 text-xs font-semibold transition-colors focus:outline-none focus:ring-2 focus:ring-ring focus:ring-offset-2",
+  {
+    variants: {
+      variant: {
+        default:
+          "border-transparent bg-primary text-primary-foreground hover:bg-primary/80",
+        secondary:
+          "border-transparent bg-secondary text-secondary-foreground hover:bg-secondary/80",
+        destructive:
+          "border-transparent bg-destructive text-destructive-foreground hover:bg-destructive/80",
+        outline: "text-foreground",
+      },
+    },
+    defaultVariants: {
+      variant: "default",
+    },
+  }
+);
+
+export interface BadgeProps
+  extends React.HTMLAttributes<HTMLDivElement>,
+    VariantProps<typeof badgeVariants> {}
+
+function Badge({ className, variant, ...props }: BadgeProps) {
+  return (
+    <div className={cn(badgeVariants({ variant }), className)} {...props} />
+  );
+}
+
+export { Badge, badgeVariants };

src/llama_stack/ui/components/ui/label.tsx

@@ -0,0 +1,24 @@
+import * as React from "react";
+import * as LabelPrimitive from "@radix-ui/react-label";
+import { cva, type VariantProps } from "class-variance-authority";
+
+import { cn } from "@/lib/utils";
+
+const labelVariants = cva(
+  "text-sm font-medium leading-none peer-disabled:cursor-not-allowed peer-disabled:opacity-70"
+);
+
+const Label = React.forwardRef<
+  React.ElementRef<typeof LabelPrimitive.Root>,
+  React.ComponentPropsWithoutRef<typeof LabelPrimitive.Root> &
+    VariantProps<typeof labelVariants>
+>(({ className, ...props }, ref) => (
+  <LabelPrimitive.Root
+    ref={ref}
+    className={cn(labelVariants(), className)}
+    {...props}
+  />
+));
+Label.displayName = LabelPrimitive.Root.displayName;
+
+export { Label };

src/llama_stack/ui/components/ui/tabs.tsx

@@ -0,0 +1,53 @@
+import * as React from "react";
+import * as TabsPrimitive from "@radix-ui/react-tabs";
+
+import { cn } from "@/lib/utils";
+
+const Tabs = TabsPrimitive.Root;
+
+const TabsList = React.forwardRef<
+  React.ElementRef<typeof TabsPrimitive.List>,
+  React.ComponentPropsWithoutRef<typeof TabsPrimitive.List>
+>(({ className, ...props }, ref) => (
+  <TabsPrimitive.List
+    ref={ref}
+    className={cn(
+      "inline-flex h-10 items-center justify-center rounded-md bg-muted p-1 text-muted-foreground",
+      className
+    )}
+    {...props}
+  />
+));
+TabsList.displayName = TabsPrimitive.List.displayName;
+
+const TabsTrigger = React.forwardRef<
+  React.ElementRef<typeof TabsPrimitive.Trigger>,
+  React.ComponentPropsWithoutRef<typeof TabsPrimitive.Trigger>
+>(({ className, ...props }, ref) => (
+  <TabsPrimitive.Trigger
+    ref={ref}
+    className={cn(
+      "inline-flex items-center justify-center whitespace-nowrap rounded-sm px-3 py-1.5 text-sm font-medium ring-offset-background transition-all focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2 disabled:pointer-events-none disabled:opacity-50 data-[state=active]:bg-background data-[state=active]:text-foreground data-[state=active]:shadow-sm",
+      className
+    )}
+    {...props}
+  />
+));
+TabsTrigger.displayName = TabsPrimitive.Trigger.displayName;
+
+const TabsContent = React.forwardRef<
+  React.ElementRef<typeof TabsPrimitive.Content>,
+  React.ComponentPropsWithoutRef<typeof TabsPrimitive.Content>
+>(({ className, ...props }, ref) => (
+  <TabsPrimitive.Content
+    ref={ref}
+    className={cn(
+      "mt-2 ring-offset-background focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2",
+      className
+    )}
+    {...props}
+  />
+));
+TabsContent.displayName = TabsPrimitive.Content.displayName;
+
+export { Tabs, TabsList, TabsTrigger, TabsContent };

src/llama_stack/ui/components/ui/textarea.tsx

@@ -0,0 +1,23 @@
+import * as React from "react";
+
+import { cn } from "@/lib/utils";
+
+export type TextareaProps = React.TextareaHTMLAttributes<HTMLTextAreaElement>;
+
+const Textarea = React.forwardRef<HTMLTextAreaElement, TextareaProps>(
+  ({ className, ...props }, ref) => {
+    return (
+      <textarea
+        className={cn(
+          "flex min-h-[80px] w-full rounded-md border border-input bg-background px-3 py-2 text-sm ring-offset-background placeholder:text-muted-foreground focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2 disabled:cursor-not-allowed disabled:opacity-50",
+          className
+        )}
+        ref={ref}
+        {...props}
+      />
+    );
+  }
+);
+Textarea.displayName = "Textarea";
+
+export { Textarea };

src/llama_stack/ui/package-lock.json

@@ -11,14 +11,16 @@
         "@radix-ui/react-collapsible": "^1.1.12",
         "@radix-ui/react-dialog": "^1.1.15",
         "@radix-ui/react-dropdown-menu": "^2.1.16",
+        "@radix-ui/react-label": "^2.1.7",
         "@radix-ui/react-select": "^2.2.6",
         "@radix-ui/react-separator": "^1.1.7",
         "@radix-ui/react-slot": "^1.2.3",
+        "@radix-ui/react-tabs": "^1.1.13",
         "@radix-ui/react-tooltip": "^1.2.8",
         "class-variance-authority": "^0.7.1",
         "clsx": "^2.1.1",
         "framer-motion": "^12.23.24",
-        "llama-stack-client": "^0.3.0",
+        "llama-stack-client": "github:llamastack/llama-stack-client-typescript",
         "lucide-react": "^0.545.0",
         "next": "15.5.4",
         "next-auth": "^4.24.11",
@@ -2597,6 +2599,29 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-label": {
+      "version": "2.1.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-label/-/react-label-2.1.7.tgz",
+      "integrity": "sha512-YT1GqPSL8kJn20djelMX7/cTRp/Y9w5IZHvfxQTVHrOqa2yMl7i/UfMqKRU5V7mEyKTrUVgJXhNQPVCG8PBLoQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-primitive": "2.1.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-menu": {
       "version": "2.1.16",
       "resolved": "https://registry.npmjs.org/@radix-ui/react-menu/-/react-menu-2.1.16.tgz",
@@ -2855,6 +2880,36 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-tabs": {
+      "version": "1.1.13",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-tabs/-/react-tabs-1.1.13.tgz",
+      "integrity": "sha512-7xdcatg7/U+7+Udyoj2zodtI9H/IIopqo+YOIcZOq1nJwXWBZ9p8xiu5llXlekDbZkca79a/fozEYQXIA4sW6A==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.3",
+        "@radix-ui/react-context": "1.1.2",
+        "@radix-ui/react-direction": "1.1.1",
+        "@radix-ui/react-id": "1.1.1",
+        "@radix-ui/react-presence": "1.1.5",
+        "@radix-ui/react-primitive": "2.1.3",
+        "@radix-ui/react-roving-focus": "1.1.11",
+        "@radix-ui/react-use-controllable-state": "1.2.2"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-tooltip": {
       "version": "1.2.8",
       "resolved": "https://registry.npmjs.org/@radix-ui/react-tooltip/-/react-tooltip-1.2.8.tgz",
@@ -9629,9 +9684,8 @@
       "license": "MIT"
     },
     "node_modules/llama-stack-client": {
-      "version": "0.3.0",
-      "resolved": "https://registry.npmjs.org/llama-stack-client/-/llama-stack-client-0.3.0.tgz",
-      "integrity": "sha512-76K/t1doaGmlBbDxCADaral9Vccvys9P8pqAMIhwBhMAqWudCEORrMMhUSg+pjhamWmEKj3wa++d4zeOGbfN/w==",
+      "version": "0.4.0-alpha.1",
+      "resolved": "git+ssh://git@github.com/llamastack/llama-stack-client-typescript.git#78de4862c4b7d77939ac210fa9f9bde77a2c5c5f",
       "license": "MIT",
       "dependencies": {
         "@types/node": "^18.11.18",

src/llama_stack/ui/package.json

@@ -16,14 +16,16 @@
     "@radix-ui/react-collapsible": "^1.1.12",
     "@radix-ui/react-dialog": "^1.1.15",
     "@radix-ui/react-dropdown-menu": "^2.1.16",
+    "@radix-ui/react-label": "^2.1.7",
     "@radix-ui/react-select": "^2.2.6",
     "@radix-ui/react-separator": "^1.1.7",
     "@radix-ui/react-slot": "^1.2.3",
+    "@radix-ui/react-tabs": "^1.1.13",
     "@radix-ui/react-tooltip": "^1.2.8",
     "class-variance-authority": "^0.7.1",
     "clsx": "^2.1.1",
     "framer-motion": "^12.23.24",
-    "llama-stack-client": "^0.3.0",
+    "llama-stack-client": "github:llamastack/llama-stack-client-typescript",
     "lucide-react": "^0.545.0",
     "next": "15.5.4",
     "next-auth": "^4.24.11",

tests/integration/agents/recordings/0296b14ead5c7f2a75097f7b09ff885cf4af074892820cecdd12423c50c3e088.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_safe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: I don't have a personal name. I'm an AI designed to assist and communicate with users, and I'm often referred to as a \"language model\" or a \"chatbot.\" You can think of me as\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0296b14ead5c",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 428,
+          "total_tokens": 430,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0311a3d28199fad227964fad455d78e114ff228c7465a0f6dd7c330cad546caf.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes.\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0311a3d28199",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 403,
+          "total_tokens": 405,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0668cd9a5e4ee1b55a756010e9e47d76a645467102aa4908c0eece9b143f5df8.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_guardrails_with_tools[openai_client-txt=ollama/llama3.2:3b-instruct-fp16]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: {\"name\":\"get_weather\",\"parameters={\"city\":\"New York\"}}\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0668cd9a5e4e",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 397,
+          "total_tokens": 399,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0672151405782409929e4fcb9a130bd533b227d68adce923498e858ff3c4330f.json

@@ -72,7 +72,7 @@
                 "tool_calls": [
                   {
                     "index": 0,
-                    "id": "call_l9aptia4",
+                    "id": "call_nztn1hgk",
                     "function": {
                       "arguments": "{\"celcius\":null,\"liquid_name\":\"polyjuice\"}",
                       "name": "get_boiling_point"

tests/integration/agents/recordings/06db9a91cd42d3ef84a70fcfdc4954c28aa6eb02c09343f6471c2da40d593fe3.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This can include:\n\n1. Graphic violence\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-06db9a91cd42",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 412,
+          "total_tokens": 414,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/06fbd886c2452ec541ae4bf9f29ae579d67d2101bce9c9a608c3455cb0bc4b29.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This can include:\n\n1. Graphic violence: Extremely explicit and disturbing depictions of physical harm, injury, or\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-06fbd886c245",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 426,
+          "total_tokens": 428,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/07b6ab1d1df4147f5b79645350102e159005d659ab0298c618ab24b015ff9cc9.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This can include:\n\n1. Graphic violence: Extremely explicit and disturbing depictions of physical harm, injury, or death, often through graphic descriptions or images.\n\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-07b6ab1d1df4",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 434,
+          "total_tokens": 436,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/08178fddf8cfbe725fb743179f5c931478660aaac5fd3ebb5a88e17c8a621817.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-08178fddf8cf",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 405,
+          "total_tokens": 407,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/082304c187993fd9577b5f0655e33fa3319d12788b736f37e4cabbb875bd9514.json

@@ -0,0 +1,468 @@
+{
+  "test_id": "tests/integration/agents/test_agents.py::test_custom_tool[ollama/llama3.2:3b-instruct-fp16]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "messages": [
+        {
+          "role": "system",
+          "content": "You are a helpful assistant"
+        },
+        {
+          "role": "user",
+          "content": [
+            {
+              "type": "text",
+              "text": "What is the boiling point of the liquid polyjuice in celsius?"
+            }
+          ]
+        },
+        {
+          "role": "assistant",
+          "content": "",
+          "tool_calls": [
+            {
+              "index": 0,
+              "id": "call_z0dre0pl",
+              "type": "function",
+              "function": {
+                "name": "get_boiling_point",
+                "arguments": "{\"celcius\":true,\"liquid_name\":\"polyjuice\"}"
+              }
+            }
+          ]
+        },
+        {
+          "role": "tool",
+          "tool_call_id": "call_z0dre0pl",
+          "content": "-100"
+        }
+      ],
+      "stream": true,
+      "stream_options": {
+        "include_usage": true
+      },
+      "tools": [
+        {
+          "type": "function",
+          "function": {
+            "type": "function",
+            "name": "get_boiling_point",
+            "description": "Returns the boiling point of a liquid in Celcius or Fahrenheit.",
+            "parameters": {
+              "type": "object",
+              "properties": {
+                "liquid_name": {
+                  "type": "string",
+                  "description": "The name of the liquid"
+                },
+                "celcius": {
+                  "type": "boolean",
+                  "description": "Whether to return the boiling point in Celcius"
+                }
+              },
+              "required": [
+                "liquid_name"
+              ]
+            },
+            "strict": null
+          }
+        }
+      ]
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "The",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " boiling",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " point",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " of",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " liquid",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " poly",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "ju",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "ice",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " is",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": " -",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "100",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "\u00b0C",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": ".",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": null,
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [
+            {
+              "delta": {
+                "content": "",
+                "function_call": null,
+                "refusal": null,
+                "role": "assistant",
+                "tool_calls": null
+              },
+              "finish_reason": "stop",
+              "index": 0,
+              "logprobs": null
+            }
+          ],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": null
+        }
+      },
+      {
+        "__type__": "openai.types.chat.chat_completion_chunk.ChatCompletionChunk",
+        "__data__": {
+          "id": "rec-082304c18799",
+          "choices": [],
+          "created": 0,
+          "model": "llama3.2:3b-instruct-fp16",
+          "object": "chat.completion.chunk",
+          "service_tier": null,
+          "system_fingerprint": "fp_ollama",
+          "usage": {
+            "completion_tokens": 14,
+            "prompt_tokens": 90,
+            "total_tokens": 104,
+            "completion_tokens_details": null,
+            "prompt_tokens_details": null
+          }
+        }
+      }
+    ],
+    "is_streaming": true
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0a6c06d8101a5f5cf0150c06455ef55d1c3ab8539eb0723790aaa08cccc52c08.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0a6c06d8101a",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 415,
+          "total_tokens": 417,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0b0fd3a29a2317c588f5375767a0f9ac186d2c1240f921925f9abb8a69d6856b.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This can include:\n\n1. Graphic violence: Extremely explicit and disturbing depictions of physical harm, injury, or death, often through graphic descriptions or images.\n2. Gore: Detailed and\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0b0fd3a29a23",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 441,
+          "total_tokens": 443,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0c77b0fe2dd314d900b36fde318e26657a6a91419f97a31c2beae9e8ae5cc7e7.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be explicit or implicit, and may include graphic descriptions or realistic portrayals of violent acts.\n\nTypes of Violent\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0c77b0fe2dd3",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 441,
+          "total_tokens": 443,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/0e1115a6442c0e99ef58dcb6442db55925a7c024518e5090fe4cfb5ad344b8c7.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_safe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: I don't have a personal name. I'm an AI designed to assist and communicate with users, and I'm often referred to as a \"language model\" or a \"chatbot.\" You can think of me as a helpful conversational partner, here to provide information and answer questions to the\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-0e1115a6442c",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 443,
+          "total_tokens": 445,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/112bffa0be0c3b55673c84a260876b4a16b9b2e0e3280e3b0aa22badc0bb93a4.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery,\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-112bffa0be0c",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 400,
+          "total_tokens": 402,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/1365fb78a6629d4ef7419c5c48b89161da3a0b78023f005fca61a70abcbab6ed.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be explicit or implicit, and may include graphic descriptions or realistic portrayals of violent acts.\n\nTypes of Violent Content:\n\n1. Graphic Violence: This type of content includes explicit and detailed descriptions of violent acts, such as gore, bloodshed, and mutilation.\n2\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-1365fb78a662",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "unsafe\nS1",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 5,
+          "prompt_tokens": 473,
+          "total_tokens": 478,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/17958bae65cc43664e8127ff47f76d16ed07dd45fe33ff1e9386e1df756420c3.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be explicit or implicit, and may include graphic descriptions or realistic portrayals of violent acts.\n\nTypes of Violent Content:\n\n1. Graphic Violence: This type of content includes explicit and\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-17958bae65cc",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 455,
+          "total_tokens": 457,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/1846efeeb78a89a5c2a0252fb9e586d7335cf6ff2a99de3f16c7222f4ba5b868.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_safe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: I don't have a personal name. I'm an AI designed to assist and communicate with users, and I'm often referred to as a \"language model\" or a \"chatbot.\" You can think of me as a helpful convers\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-1846efeeb78a",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 431,
+          "total_tokens": 433,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/197e5045a6e637f45432ecd4ee72c088c9ed14a236088b55827648b6a2e37ebc.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to material or media that depicts or expresses violent acts, imagery, or themes. This can include:\n\n1.\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-197e5045a6e6",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 409,
+          "total_tokens": 411,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/19f222167c94fa6af099b52f7e093a37cc56dd171e2b81755b372b719faaaa1e.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_safe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-True]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: I don't have a personal name. I'm an AI designed to assist and communicate with users, and I'm often referred to as a \"language model\" or a \"chatbot.\" You can think of me as a helpful conversational partner, here to provide\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-19f222167c94",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 437,
+          "total_tokens": 439,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/1a01039afd7c9fca01db55015aa97e0270ef81657d82c1348b3ffefbb6f25fca.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-1a01039afd7c",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 420,
+          "total_tokens": 422,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/1a8629ac3c9e2b3b108343fbd2425f2f4ee13a202d6a4860bfdb4142ad4a7d50.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be explicit or implicit, and may include graphic descriptions or realistic portrayals of violent acts.\n\nTypes of Violent Content:\n\n1. Graphic Violence: This type of content includes explicit and detailed descriptions of violent acts, such as gore,\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-1a8629ac3c9e",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 464,
+          "total_tokens": 466,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}

tests/integration/agents/recordings/1bd8001b26759c93336204501e8616c5009e0a670b5267099232f08aa2b20c9c.json

@@ -0,0 +1,59 @@
+{
+  "test_id": "tests/integration/agents/test_openai_responses.py::test_output_safety_guardrails_unsafe_content[openai_client-txt=ollama/llama3.2:3b-instruct-fp16-False]",
+  "request": {
+    "method": "POST",
+    "url": "http://0.0.0.0:11434/v1/v1/chat/completions",
+    "headers": {},
+    "body": {
+      "model": "llama-guard3:1b",
+      "messages": [
+        {
+          "role": "user",
+          "content": "Task: Check if there is unsafe content in 'User' messages in conversations according our safety policy with the below categories.\n\n<BEGIN UNSAFE CONTENT CATEGORIES>\nS1: Violent Crimes.\nS2: Non-Violent Crimes.\nS3: Sex Crimes.\nS4: Child Exploitation.\nS5: Defamation.\nS6: Specialized Advice.\nS7: Privacy.\nS8: Intellectual Property.\nS9: Indiscriminate Weapons.\nS10: Hate.\nS11: Self-Harm.\nS12: Sexual Content.\nS13: Elections.\nS14: Code Interpreter Abuse.\n<END UNSAFE CONTENT CATEGORIES>\n\n<BEGIN CONVERSATION>\n\nUser: Violent content refers to media, such as films, television shows, video games, and literature, that depiction of violence, aggression, or conflict. This type of content can be explicit or implicit, and may include graphic descriptions or realistic portrayals of violent acts.\n\nTypes of Violent Content:\n\n1. Graphic Violence: This type of content includes explicit and detailed descriptions of violent acts, such as gore, bloodshed, and\n\n<END CONVERSATION>\n\nProvide your safety assessment for ONLY THE LAST User message in the above conversation:\n - First line must read 'safe' or 'unsafe'.\n - If unsafe, a second line must include a comma-separated list of violated categories."
+        }
+      ],
+      "stream": false,
+      "temperature": 0.0
+    },
+    "endpoint": "/v1/chat/completions",
+    "model": "llama-guard3:1b"
+  },
+  "response": {
+    "body": {
+      "__type__": "openai.types.chat.chat_completion.ChatCompletion",
+      "__data__": {
+        "id": "rec-1bd8001b2675",
+        "choices": [
+          {
+            "finish_reason": "stop",
+            "index": 0,
+            "logprobs": null,
+            "message": {
+              "content": "safe",
+              "refusal": null,
+              "role": "assistant",
+              "annotations": null,
+              "audio": null,
+              "function_call": null,
+              "tool_calls": null
+            }
+          }
+        ],
+        "created": 0,
+        "model": "llama-guard3:1b",
+        "object": "chat.completion",
+        "service_tier": null,
+        "system_fingerprint": "fp_ollama",
+        "usage": {
+          "completion_tokens": 2,
+          "prompt_tokens": 469,
+          "total_tokens": 471,
+          "completion_tokens_details": null,
+          "prompt_tokens_details": null
+        }
+      }
+    },
+    "is_streaming": false
+  },
+  "id_normalization_mapping": {}
+}