chore(package): migrate to src/ layout (#3920)

Migrates package structure to src/ layout following Python packaging
best practices.

All code moved from `llama_stack/` to `src/llama_stack/`. Public API
unchanged - imports remain `import llama_stack.*`.

Updated build configs, pre-commit hooks, scripts, and GitHub workflows
accordingly. All hooks pass, package builds cleanly.

**Developer note**: Reinstall after pulling: `pip install -e .`

.github/workflows/integration-auth-tests.yml

@@ -9,8 +9,8 @@ on:
     branches: [ main ]
     paths:
       - 'distributions/**'
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
-      - '!llama_stack/ui/**'
+      - '!src/llama_stack/ui/**'
       - 'tests/integration/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/integration-sql-store-tests.yml

@@ -8,7 +8,7 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/providers/utils/sqlstore/**'
+      - 'src/llama_stack/providers/utils/sqlstore/**'
       - 'tests/integration/sqlstore/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/integration-tests.yml

@@ -9,8 +9,8 @@ on:
     branches: [ main ]
     types: [opened, synchronize, reopened]
     paths:
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
-      - '!llama_stack/ui/**'
+      - '!src/llama_stack/ui/**'
       - 'tests/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/integration-vector-io-tests.yml

@@ -8,8 +8,8 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
-      - '!llama_stack/ui/**'
+      - '!src/llama_stack/ui/**'
       - 'tests/integration/vector_io/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/pre-commit.yml

@@ -41,11 +41,11 @@ jobs:
         with:
           node-version: '20'
           cache: 'npm'
-          cache-dependency-path: 'llama_stack/ui/'
+          cache-dependency-path: 'src/llama_stack/ui/'
 
       - name: Install npm dependencies
         run: npm ci
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
 
       - name: Run pre-commit
         id: precommit

.github/workflows/precommit-trigger.yml

@@ -145,12 +145,12 @@ jobs:
         with:
           node-version: '20'
           cache: 'npm'
-          cache-dependency-path: 'llama_stack/ui/'
+          cache-dependency-path: 'src/llama_stack/ui/'
 
       - name: Install npm dependencies
         if: steps.check_author.outputs.authorized == 'true'
         run: npm ci
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
 
       - name: Run pre-commit
         if: steps.check_author.outputs.authorized == 'true'

.github/workflows/providers-build.yml

@@ -7,24 +7,24 @@ on:
     branches:
       - main
     paths:
-      - 'llama_stack/cli/stack/build.py'
+      - 'src/llama_stack/cli/stack/build.py'
-      - 'llama_stack/cli/stack/_build.py'
+      - 'src/llama_stack/cli/stack/_build.py'
-      - 'llama_stack/core/build.*'
+      - 'src/llama_stack/core/build.*'
-      - 'llama_stack/core/*.sh'
+      - 'src/llama_stack/core/*.sh'
       - '.github/workflows/providers-build.yml'
-      - 'llama_stack/distributions/**'
+      - 'src/llama_stack/distributions/**'
       - 'pyproject.toml'
       - 'containers/Containerfile'
       - '.dockerignore'
 
   pull_request:
     paths:
-      - 'llama_stack/cli/stack/build.py'
+      - 'src/llama_stack/cli/stack/build.py'
-      - 'llama_stack/cli/stack/_build.py'
+      - 'src/llama_stack/cli/stack/_build.py'
-      - 'llama_stack/core/build.*'
+      - 'src/llama_stack/core/build.*'
-      - 'llama_stack/core/*.sh'
+      - 'src/llama_stack/core/*.sh'
       - '.github/workflows/providers-build.yml'
-      - 'llama_stack/distributions/**'
+      - 'src/llama_stack/distributions/**'
       - 'pyproject.toml'
       - 'containers/Containerfile'
       - '.dockerignore'
@@ -45,7 +45,7 @@ jobs:
       - name: Generate Distribution List
         id: set-matrix
         run: |
-          distros=$(ls llama_stack/distributions/*/*build.yaml | awk -F'/' '{print $(NF-1)}' | jq -R -s -c 'split("\n")[:-1]')
+          distros=$(ls src/llama_stack/distributions/*/*build.yaml | awk -F'/' '{print $(NF-1)}' | jq -R -s -c 'split("\n")[:-1]')
           echo "distros=$distros" >> "$GITHUB_OUTPUT"
 
   build:
@@ -107,13 +107,13 @@ jobs:
 
       - name: Build container image
         run: |
-          BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "python:3.12-slim"' llama_stack/distributions/ci-tests/build.yaml)
+          BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "python:3.12-slim"' src/llama_stack/distributions/ci-tests/build.yaml)
           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/llama_stack/distributions/ci-tests/run.yaml \
+            --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml \
             -t llama-stack:ci-tests
 
       - name: Inspect the container image entrypoint
@@ -143,17 +143,17 @@ jobs:
         run: |
           yq -i '
             .distribution_spec.container_image = "registry.access.redhat.com/ubi9:latest"
-          ' llama_stack/distributions/ci-tests/build.yaml
+          ' src/llama_stack/distributions/ci-tests/build.yaml
 
       - name: Build UBI9 container image
         run: |
-          BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "registry.access.redhat.com/ubi9:latest"' llama_stack/distributions/ci-tests/build.yaml)
+          BASE_IMAGE=$(yq -r '.distribution_spec.container_image // "registry.access.redhat.com/ubi9:latest"' src/llama_stack/distributions/ci-tests/build.yaml)
           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/llama_stack/distributions/ci-tests/run.yaml \
+            --build-arg RUN_CONFIG_PATH=/workspace/src/llama_stack/distributions/ci-tests/run.yaml \
             -t llama-stack:ci-tests-ubi9
 
       - name: Inspect UBI9 image

.github/workflows/providers-list-deps.yml

@@ -7,22 +7,22 @@ on:
     branches:
       - main
     paths:
-      - 'llama_stack/cli/stack/list_deps.py'
+      - 'src/llama_stack/cli/stack/list_deps.py'
-      - 'llama_stack/cli/stack/_list_deps.py'
+      - 'src/llama_stack/cli/stack/_list_deps.py'
-      - 'llama_stack/core/build.*'
+      - 'src/llama_stack/core/build.*'
-      - 'llama_stack/core/*.sh'
+      - 'src/llama_stack/core/*.sh'
       - '.github/workflows/providers-list-deps.yml'
-      - 'llama_stack/templates/**'
+      - 'src/llama_stack/templates/**'
       - 'pyproject.toml'
 
   pull_request:
     paths:
-      - 'llama_stack/cli/stack/list_deps.py'
+      - 'src/llama_stack/cli/stack/list_deps.py'
-      - 'llama_stack/cli/stack/_list_deps.py'
+      - 'src/llama_stack/cli/stack/_list_deps.py'
-      - 'llama_stack/core/build.*'
+      - 'src/llama_stack/core/build.*'
-      - 'llama_stack/core/*.sh'
+      - 'src/llama_stack/core/*.sh'
       - '.github/workflows/providers-list-deps.yml'
-      - 'llama_stack/templates/**'
+      - 'src/llama_stack/templates/**'
       - 'pyproject.toml'
 
 concurrency:
@@ -41,7 +41,7 @@ jobs:
       - name: Generate Distribution List
         id: set-matrix
         run: |
-          distros=$(ls llama_stack/distributions/*/*build.yaml | awk -F'/' '{print $(NF-1)}' | jq -R -s -c 'split("\n")[:-1]')
+          distros=$(ls src/llama_stack/distributions/*/*build.yaml | awk -F'/' '{print $(NF-1)}' | jq -R -s -c 'split("\n")[:-1]')
           echo "distros=$distros" >> "$GITHUB_OUTPUT"
 
   list-deps:
@@ -102,4 +102,4 @@ jobs:
           USE_COPY_NOT_MOUNT: "true"
           LLAMA_STACK_DIR: "."
         run: |
-          uv run llama stack list-deps llama_stack/distributions/ci-tests/build.yaml
+          uv run llama stack list-deps src/llama_stack/distributions/ci-tests/build.yaml

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

@@ -10,7 +10,7 @@ on:
     branches:
       - main
     paths-ignore:
-        - 'llama_stack/ui/**'
+        - 'src/llama_stack/ui/**'
 
 jobs:
   build:

.github/workflows/test-external-provider-module.yml

@@ -8,7 +8,7 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
       - 'tests/integration/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/test-external.yml

@@ -8,8 +8,8 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
-      - '!llama_stack/ui/**'
+      - '!src/llama_stack/ui/**'
       - 'tests/integration/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/ui-unit-tests.yml

@@ -8,7 +8,7 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/ui/**'
+      - 'src/llama_stack/ui/**'
       - '.github/workflows/ui-unit-tests.yml' # This workflow
   workflow_dispatch:
 
@@ -33,22 +33,22 @@ jobs:
         with:
           node-version: ${{ matrix.node-version }}
           cache: 'npm'
-          cache-dependency-path: 'llama_stack/ui/package-lock.json'
+          cache-dependency-path: 'src/llama_stack/ui/package-lock.json'
 
       - name: Install dependencies
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
         run: npm ci
 
       - name: Run linting
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
         run: npm run lint
 
       - name: Run format check
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
         run: npm run format:check
 
       - name: Run unit tests
-        working-directory: llama_stack/ui
+        working-directory: src/llama_stack/ui
         env:
           CI: true
 

.github/workflows/unit-tests.yml

@@ -8,8 +8,8 @@ on:
   pull_request:
     branches: [ main ]
     paths:
-      - 'llama_stack/**'
+      - 'src/llama_stack/**'
-      - '!llama_stack/ui/**'
+      - '!src/llama_stack/ui/**'
       - 'tests/unit/**'
       - 'uv.lock'
       - 'pyproject.toml'

.pre-commit-config.yaml

@@ -42,7 +42,7 @@ repos:
     hooks:
     -   id: ruff
         args: [ --fix ]
-        exclude: ^llama_stack/strong_typing/.*$
+        exclude: ^src/llama_stack/strong_typing/.*$
     -   id: ruff-format
 
 -   repo: https://github.com/adamchainz/blacken-docs
@@ -86,7 +86,7 @@ repos:
         language: python
         pass_filenames: false
         require_serial: true
-        files: ^llama_stack/distributions/.*$|^llama_stack/providers/.*/inference/.*/models\.py$
+        files: ^src/llama_stack/distributions/.*$|^src/llama_stack/providers/.*/inference/.*/models\.py$
       - id: provider-codegen
         name: Provider Codegen
         additional_dependencies:
@@ -95,7 +95,7 @@ repos:
         language: python
         pass_filenames: false
         require_serial: true
-        files: ^llama_stack/providers/.*$
+        files: ^src/llama_stack/providers/.*$
       - id: openapi-codegen
         name: API Spec Codegen
         additional_dependencies:
@@ -104,7 +104,7 @@ repos:
         language: python
         pass_filenames: false
         require_serial: true
-        files: ^llama_stack/apis/|^docs/openapi_generator/
+        files: ^src/llama_stack/apis/|^docs/openapi_generator/
       - id: check-workflows-use-hashes
         name: Check GitHub Actions use SHA-pinned actions
         entry: ./scripts/check-workflows-use-hashes.sh
@@ -120,7 +120,7 @@ repos:
         pass_filenames: false
         require_serial: true
         always_run: true
-        files: ^llama_stack/.*$
+        files: ^src/llama_stack/.*$
       - id: forbid-pytest-asyncio
         name: Block @pytest.mark.asyncio and @pytest_asyncio.fixture
         entry: bash
@@ -150,7 +150,7 @@ repos:
         name: Format & Lint UI
         entry: bash ./scripts/run-ui-linter.sh
         language: system
-        files: ^llama_stack/ui/.*\.(ts|tsx)$
+        files: ^src/llama_stack/ui/.*\.(ts|tsx)$
         pass_filenames: false
         require_serial: true
 

MANIFEST.in

@@ -1,11 +1,11 @@
 include pyproject.toml
-include llama_stack/models/llama/llama3/tokenizer.model
+include src/llama_stack/models/llama/llama3/tokenizer.model
-include llama_stack/models/llama/llama4/tokenizer.model
+include src/llama_stack/models/llama/llama4/tokenizer.model
-include llama_stack/core/*.sh
+include src/llama_stack/core/*.sh
-include llama_stack/cli/scripts/*.sh
+include src/llama_stack/cli/scripts/*.sh
-include llama_stack/distributions/*/*.yaml
+include src/llama_stack/distributions/*/*.yaml
-exclude llama_stack/distributions/ci-tests
+exclude src/llama_stack/distributions/ci-tests
 include tests/integration/test_cases/inference/*.json
-include llama_stack/models/llama/*/*.md
+include src/llama_stack/models/llama/*/*.md
-include llama_stack/tests/integration/*.jpg
+include src/llama_stack/tests/integration/*.jpg
-prune llama_stack/distributions/ci-tests
+prune src/llama_stack/distributions/ci-tests

pyproject.toml

@@ -150,7 +150,7 @@ llama = "llama_stack.cli.llama:main"
 install-wheel-from-presigned = "llama_stack.cli.scripts.run:install_wheel_from_presigned"
 
 [tool.setuptools.packages.find]
-where = ["."]
+where = ["src"]
 include = ["llama_stack", "llama_stack.*"]
 
 [[tool.uv.index]]
@@ -217,17 +217,17 @@ unfixable = [
 # Ignore the following errors for the following files
 [tool.ruff.lint.per-file-ignores]
 "tests/**/*.py" = ["DTZ"] # Ignore datetime rules for tests
-"llama_stack/providers/inline/scoring/basic/utils/ifeval_utils.py" = ["RUF001"]
+"src/llama_stack/providers/inline/scoring/basic/utils/ifeval_utils.py" = ["RUF001"]
-"llama_stack/providers/inline/scoring/basic/scoring_fn/fn_defs/regex_parser_multiple_choice_answer.py" = [
+"src/llama_stack/providers/inline/scoring/basic/scoring_fn/fn_defs/regex_parser_multiple_choice_answer.py" = [
     "RUF001",
     "PLE2515",
 ]
-"llama_stack/apis/**/__init__.py" = [
+"src/llama_stack/apis/**/__init__.py" = [
     "F403",
 ] # Using import * is acceptable (or at least tolerated) in an __init__.py of a package API
 
 [tool.mypy]
-mypy_path = ["llama_stack"]
+mypy_path = ["src"]
 packages = ["llama_stack"]
 plugins = ['pydantic.mypy']
 disable_error_code = []
@@ -239,77 +239,77 @@ follow_imports = "silent"
 # to exclude the entire directory.
 exclude = [
     # As we fix more and more of these, we should remove them from the list
-    "^llama_stack.core/build\\.py$",
+    "^src/llama_stack/core/build\\.py$",
-    "^llama_stack.core/client\\.py$",
+    "^src/llama_stack/core/client\\.py$",
-    "^llama_stack.core/request_headers\\.py$",
+    "^src/llama_stack/core/request_headers\\.py$",
-    "^llama_stack.core/routers/",
+    "^src/llama_stack/core/routers/",
-    "^llama_stack.core/routing_tables/",
+    "^src/llama_stack/core/routing_tables/",
-    "^llama_stack.core/server/endpoints\\.py$",
+    "^src/llama_stack/core/server/endpoints\\.py$",
-    "^llama_stack.core/server/server\\.py$",
+    "^src/llama_stack/core/server/server\\.py$",
-    "^llama_stack.core/stack\\.py$",
+    "^src/llama_stack/core/stack\\.py$",
-    "^llama_stack.core/store/registry\\.py$",
+    "^src/llama_stack/core/store/registry\\.py$",
-    "^llama_stack.core/utils/exec\\.py$",
+    "^src/llama_stack/core/utils/exec\\.py$",
-    "^llama_stack.core/utils/prompt_for_config\\.py$",
+    "^src/llama_stack/core/utils/prompt_for_config\\.py$",
-    "^llama_stack/models/llama/llama3/interface\\.py$",
+    "^src/llama_stack/models/llama/llama3/interface\\.py$",
-    "^llama_stack/models/llama/llama3/tokenizer\\.py$",
+    "^src/llama_stack/models/llama/llama3/tokenizer\\.py$",
-    "^llama_stack/models/llama/llama3/tool_utils\\.py$",
+    "^src/llama_stack/models/llama/llama3/tool_utils\\.py$",
-    "^llama_stack/providers/inline/agents/meta_reference/",
+    "^src/llama_stack/providers/inline/agents/meta_reference/",
-    "^llama_stack/providers/inline/datasetio/localfs/",
+    "^src/llama_stack/providers/inline/datasetio/localfs/",
-    "^llama_stack/providers/inline/eval/meta_reference/eval\\.py$",
+    "^src/llama_stack/providers/inline/eval/meta_reference/eval\\.py$",
-    "^llama_stack/providers/inline/inference/meta_reference/inference\\.py$",
+    "^src/llama_stack/providers/inline/inference/meta_reference/inference\\.py$",
-    "^llama_stack/models/llama/llama3/generation\\.py$",
+    "^src/llama_stack/models/llama/llama3/generation\\.py$",
-    "^llama_stack/models/llama/llama3/multimodal/model\\.py$",
+    "^src/llama_stack/models/llama/llama3/multimodal/model\\.py$",
-    "^llama_stack/models/llama/llama4/",
+    "^src/llama_stack/models/llama/llama4/",
-    "^llama_stack/providers/inline/inference/sentence_transformers/sentence_transformers\\.py$",
+    "^src/llama_stack/providers/inline/inference/sentence_transformers/sentence_transformers\\.py$",
-    "^llama_stack/providers/inline/post_training/common/validator\\.py$",
+    "^src/llama_stack/providers/inline/post_training/common/validator\\.py$",
-    "^llama_stack/providers/inline/safety/code_scanner/",
+    "^src/llama_stack/providers/inline/safety/code_scanner/",
-    "^llama_stack/providers/inline/safety/llama_guard/",
+    "^src/llama_stack/providers/inline/safety/llama_guard/",
-    "^llama_stack/providers/inline/scoring/basic/",
+    "^src/llama_stack/providers/inline/scoring/basic/",
-    "^llama_stack/providers/inline/scoring/braintrust/",
+    "^src/llama_stack/providers/inline/scoring/braintrust/",
-    "^llama_stack/providers/inline/scoring/llm_as_judge/",
+    "^src/llama_stack/providers/inline/scoring/llm_as_judge/",
-    "^llama_stack/providers/remote/agents/sample/",
+    "^src/llama_stack/providers/remote/agents/sample/",
-    "^llama_stack/providers/remote/datasetio/huggingface/",
+    "^src/llama_stack/providers/remote/datasetio/huggingface/",
-    "^llama_stack/providers/remote/datasetio/nvidia/",
+    "^src/llama_stack/providers/remote/datasetio/nvidia/",
-    "^llama_stack/providers/remote/inference/bedrock/",
+    "^src/llama_stack/providers/remote/inference/bedrock/",
-    "^llama_stack/providers/remote/inference/nvidia/",
+    "^src/llama_stack/providers/remote/inference/nvidia/",
-    "^llama_stack/providers/remote/inference/passthrough/",
+    "^src/llama_stack/providers/remote/inference/passthrough/",
-    "^llama_stack/providers/remote/inference/runpod/",
+    "^src/llama_stack/providers/remote/inference/runpod/",
-    "^llama_stack/providers/remote/inference/tgi/",
+    "^src/llama_stack/providers/remote/inference/tgi/",
-    "^llama_stack/providers/remote/inference/watsonx/",
+    "^src/llama_stack/providers/remote/inference/watsonx/",
-    "^llama_stack/providers/remote/safety/bedrock/",
+    "^src/llama_stack/providers/remote/safety/bedrock/",
-    "^llama_stack/providers/remote/safety/nvidia/",
+    "^src/llama_stack/providers/remote/safety/nvidia/",
-    "^llama_stack/providers/remote/safety/sambanova/",
+    "^src/llama_stack/providers/remote/safety/sambanova/",
-    "^llama_stack/providers/remote/safety/sample/",
+    "^src/llama_stack/providers/remote/safety/sample/",
-    "^llama_stack/providers/remote/tool_runtime/bing_search/",
+    "^src/llama_stack/providers/remote/tool_runtime/bing_search/",
-    "^llama_stack/providers/remote/tool_runtime/brave_search/",
+    "^src/llama_stack/providers/remote/tool_runtime/brave_search/",
-    "^llama_stack/providers/remote/tool_runtime/model_context_protocol/",
+    "^src/llama_stack/providers/remote/tool_runtime/model_context_protocol/",
-    "^llama_stack/providers/remote/tool_runtime/tavily_search/",
+    "^src/llama_stack/providers/remote/tool_runtime/tavily_search/",
-    "^llama_stack/providers/remote/tool_runtime/wolfram_alpha/",
+    "^src/llama_stack/providers/remote/tool_runtime/wolfram_alpha/",
-    "^llama_stack/providers/remote/post_training/nvidia/",
+    "^src/llama_stack/providers/remote/post_training/nvidia/",
-    "^llama_stack/providers/remote/vector_io/chroma/",
+    "^src/llama_stack/providers/remote/vector_io/chroma/",
-    "^llama_stack/providers/remote/vector_io/milvus/",
+    "^src/llama_stack/providers/remote/vector_io/milvus/",
-    "^llama_stack/providers/remote/vector_io/pgvector/",
+    "^src/llama_stack/providers/remote/vector_io/pgvector/",
-    "^llama_stack/providers/remote/vector_io/qdrant/",
+    "^src/llama_stack/providers/remote/vector_io/qdrant/",
-    "^llama_stack/providers/remote/vector_io/sample/",
+    "^src/llama_stack/providers/remote/vector_io/sample/",
-    "^llama_stack/providers/remote/vector_io/weaviate/",
+    "^src/llama_stack/providers/remote/vector_io/weaviate/",
-    "^llama_stack/providers/utils/bedrock/client\\.py$",
+    "^src/llama_stack/providers/utils/bedrock/client\\.py$",
-    "^llama_stack/providers/utils/bedrock/refreshable_boto_session\\.py$",
+    "^src/llama_stack/providers/utils/bedrock/refreshable_boto_session\\.py$",
-    "^llama_stack/providers/utils/inference/embedding_mixin\\.py$",
+    "^src/llama_stack/providers/utils/inference/embedding_mixin\\.py$",
-    "^llama_stack/providers/utils/inference/litellm_openai_mixin\\.py$",
+    "^src/llama_stack/providers/utils/inference/litellm_openai_mixin\\.py$",
-    "^llama_stack/providers/utils/inference/model_registry\\.py$",
+    "^src/llama_stack/providers/utils/inference/model_registry\\.py$",
-    "^llama_stack/providers/utils/inference/openai_compat\\.py$",
+    "^src/llama_stack/providers/utils/inference/openai_compat\\.py$",
-    "^llama_stack/providers/utils/inference/prompt_adapter\\.py$",
+    "^src/llama_stack/providers/utils/inference/prompt_adapter\\.py$",
-    "^llama_stack/providers/utils/kvstore/kvstore\\.py$",
+    "^src/llama_stack/providers/utils/kvstore/kvstore\\.py$",
-    "^llama_stack/providers/utils/kvstore/postgres/postgres\\.py$",
+    "^src/llama_stack/providers/utils/kvstore/postgres/postgres\\.py$",
-    "^llama_stack/providers/utils/kvstore/redis/redis\\.py$",
+    "^src/llama_stack/providers/utils/kvstore/redis/redis\\.py$",
-    "^llama_stack/providers/utils/memory/vector_store\\.py$",
+    "^src/llama_stack/providers/utils/memory/vector_store\\.py$",
-    "^llama_stack/providers/utils/scoring/aggregation_utils\\.py$",
+    "^src/llama_stack/providers/utils/scoring/aggregation_utils\\.py$",
-    "^llama_stack/providers/utils/scoring/base_scoring_fn\\.py$",
+    "^src/llama_stack/providers/utils/scoring/base_scoring_fn\\.py$",
-    "^llama_stack/providers/utils/telemetry/dataset_mixin\\.py$",
+    "^src/llama_stack/providers/utils/telemetry/dataset_mixin\\.py$",
-    "^llama_stack/providers/utils/telemetry/trace_protocol\\.py$",
+    "^src/llama_stack/providers/utils/telemetry/trace_protocol\\.py$",
-    "^llama_stack/providers/utils/telemetry/tracing\\.py$",
+    "^src/llama_stack/providers/utils/telemetry/tracing\\.py$",
-    "^llama_stack/strong_typing/auxiliary\\.py$",
+    "^src/llama_stack/strong_typing/auxiliary\\.py$",
-    "^llama_stack/distributions/template\\.py$",
+    "^src/llama_stack/distributions/template\\.py$",
 ]
 
 [[tool.mypy.overrides]]

scripts/check-init-py.sh

@@ -16,7 +16,7 @@ if (( BASH_VERSINFO[0] < 4 )); then
     exit 1
 fi
 
-PACKAGE_DIR="${1:-llama_stack}"
+PACKAGE_DIR="${1:-src/llama_stack}"
 
 if [ ! -d "$PACKAGE_DIR" ]; then
     echo "ERROR: Package directory '$PACKAGE_DIR' does not exist"

scripts/distro_codegen.py

@@ -93,7 +93,7 @@ def pre_import_distros(distro_dirs: list[Path]) -> None:
 
 
 def main():
-    distros_dir = REPO_ROOT / "llama_stack" / "distributions"
+    distros_dir = REPO_ROOT / "src" / "llama_stack" / "distributions"
     change_tracker = ChangedPathTracker()
 
     with Progress(

scripts/run-ui-linter.sh

@@ -6,7 +6,7 @@
 # the root directory of this source tree.
 
 set -e
-cd llama_stack/ui
+cd src/llama_stack/ui
 
 if [ ! -d node_modules ] || [ ! -x node_modules/.bin/prettier ] || [ ! -x node_modules/.bin/eslint ]; then
   echo "UI dependencies not installed, skipping prettier/linter check"

scripts/unit-tests.sh

@@ -27,4 +27,4 @@ fi
 
 # Run unit tests with coverage
 uv run --python "$PYTHON_VERSION" --with-editable . --group unit \
-    coverage run --source=llama_stack -m pytest -s -v tests/unit/ "$@"
+    coverage run --source=src/llama_stack -m pytest -s -v tests/unit/ "$@"