Merge remote-tracking branch 'upstream/main' into strip-telem

.coveragerc

@@ -5,7 +5,7 @@ omit =
     */llama_stack/templates/*
     .venv/*
     */llama_stack/cli/scripts/*
-    */llama_stack/ui/*
+    */llama_stack_ui/*
     */llama_stack/distribution/ui/*
     */llama_stack/strong_typing/*
     */llama_stack/env.py

.github/dependabot.yml

@@ -22,7 +22,7 @@ updates:
       prefix: chore(python-deps)
 
   - package-ecosystem: npm
-    directory: "/llama_stack/ui"
+    directory: "/llama_stack_ui"
     schedule:
       interval: "weekly"
       day: "saturday"

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

@@ -14,7 +14,7 @@ on:
     paths:
       - 'distributions/**'
       - 'src/llama_stack/**'
-      - '!src/llama_stack/ui/**'
+      - '!src/llama_stack_ui/**'
       - 'tests/integration/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/integration-tests.yml

@@ -14,7 +14,7 @@ on:
     types: [opened, synchronize, reopened]
     paths:
       - 'src/llama_stack/**'
-      - '!src/llama_stack/ui/**'
+      - '!src/llama_stack_ui/**'
       - 'tests/**'
       - 'uv.lock'
       - 'pyproject.toml'

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

@@ -13,7 +13,7 @@ on:
       - 'release-[0-9]+.[0-9]+.x'
     paths:
       - 'src/llama_stack/**'
-      - '!src/llama_stack/ui/**'
+      - '!src/llama_stack_ui/**'
       - 'tests/integration/vector_io/**'
       - 'uv.lock'
       - 'pyproject.toml'

.github/workflows/pre-commit.yml

@@ -43,14 +43,14 @@ jobs:
         with:
           node-version: '20'
           cache: 'npm'
-          cache-dependency-path: 'src/llama_stack/ui/'
+          cache-dependency-path: 'src/llama_stack_ui/'
 
       - name: Set up uv
         uses: astral-sh/setup-uv@85856786d1ce8acfbcc2f13a5f3fbd6b938f9f41 # v7.1.2
 
       - name: Install npm dependencies
         run: npm ci
-        working-directory: src/llama_stack/ui
+        working-directory: src/llama_stack_ui
 
       - name: Install pre-commit
         run: python -m pip install pre-commit

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

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

.github/workflows/test-external.yml

@@ -9,7 +9,7 @@ on:
     branches: [ main ]
     paths:
       - 'src/llama_stack/**'
-      - '!src/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:
-      - 'src/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: 'src/llama_stack/ui/package-lock.json'
+          cache-dependency-path: 'src/llama_stack_ui/package-lock.json'
 
       - name: Install dependencies
-        working-directory: src/llama_stack/ui
+        working-directory: src/llama_stack_ui
         run: npm ci
 
       - name: Run linting
-        working-directory: src/llama_stack/ui
+        working-directory: src/llama_stack_ui
         run: npm run lint
 
       - name: Run format check
-        working-directory: src/llama_stack/ui
+        working-directory: src/llama_stack_ui
         run: npm run format:check
 
       - name: Run unit tests
-        working-directory: src/llama_stack/ui
+        working-directory: src/llama_stack_ui
         env:
           CI: true
 

.github/workflows/unit-tests.yml

@@ -13,7 +13,7 @@ on:
       - 'release-[0-9]+.[0-9]+.x'
     paths:
       - 'src/llama_stack/**'
-      - '!src/llama_stack/ui/**'
+      - '!src/llama_stack_ui/**'
       - 'tests/unit/**'
       - 'uv.lock'
       - 'pyproject.toml'

.pre-commit-config.yaml

@@ -161,7 +161,7 @@ repos:
         name: Format & Lint UI
         entry: bash ./scripts/run-ui-linter.sh
         language: system
-        files: ^src/llama_stack/ui/.*\.(ts|tsx)$
+        files: ^src/llama_stack_ui/.*\.(ts|tsx)$
         pass_filenames: false
         require_serial: true
 

client-sdks/stainless/config-not-source-of-truth-yet.yml

@@ -0,0 +1,525 @@
+# yaml-language-server: $schema=https://app.stainlessapi.com/config-internal.schema.json
+
+organization:
+  # Name of your organization or company, used to determine the name of the client
+  # and headings.
+  name: llama-stack-client
+  docs: https://llama-stack.readthedocs.io/en/latest/
+  contact: llamastack@meta.com
+security:
+  - {}
+  - BearerAuth: []
+security_schemes:
+  BearerAuth:
+    type: http
+    scheme: bearer
+# `targets` define the output targets and their customization options, such as
+# whether to emit the Node SDK and what it's package name should be.
+targets:
+  node:
+    package_name: llama-stack-client
+    production_repo: llamastack/llama-stack-client-typescript
+    publish:
+      npm: false
+  python:
+    package_name: llama_stack_client
+    production_repo: llamastack/llama-stack-client-python
+    options:
+      use_uv: true
+    publish:
+      pypi: true
+    project_name: llama_stack_client
+  kotlin:
+    reverse_domain: com.llama_stack_client.api
+    production_repo: null
+    publish:
+      maven: false
+  go:
+    package_name: llama-stack-client
+    production_repo: llamastack/llama-stack-client-go
+    options:
+      enable_v2: true
+      back_compat_use_shared_package: false
+
+# `client_settings` define settings for the API client, such as extra constructor
+# arguments (used for authentication), retry behavior, idempotency, etc.
+client_settings:
+  default_env_prefix: LLAMA_STACK_CLIENT
+  opts:
+    api_key:
+      type: string
+      read_env: LLAMA_STACK_CLIENT_API_KEY
+      auth: { security_scheme: BearerAuth }
+      nullable: true
+
+# `environments` are a map of the name of the environment (e.g. "sandbox",
+# "production") to the corresponding url to use.
+environments:
+  production: http://any-hosted-llama-stack.com
+
+# `pagination` defines [pagination schemes] which provides a template to match
+# endpoints and generate next-page and auto-pagination helpers in the SDKs.
+pagination:
+  - name: datasets_iterrows
+    type: offset
+    request:
+      dataset_id:
+        type: string
+      start_index:
+        type: integer
+        x-stainless-pagination-property:
+          purpose: offset_count_param
+      limit:
+        type: integer
+    response:
+      data:
+        type: array
+        items:
+          type: object
+      next_index:
+        type: integer
+        x-stainless-pagination-property:
+          purpose: offset_count_start_field
+  - name: openai_cursor_page
+    type: cursor
+    request:
+      limit:
+        type: integer
+      after:
+        type: string
+        x-stainless-pagination-property:
+          purpose: next_cursor_param
+    response:
+      data:
+        type: array
+        items: {}
+      has_more:
+        type: boolean
+      last_id:
+        type: string
+        x-stainless-pagination-property:
+          purpose: next_cursor_field
+# `resources` define the structure and organziation for your API, such as how
+# methods and models are grouped together and accessed. See the [configuration
+# guide] for more information.
+#
+# [configuration guide]:
+#   https://app.stainlessapi.com/docs/guides/configure#resources
+resources:
+  $shared:
+    models:
+      interleaved_content_item: InterleavedContentItem
+      interleaved_content: InterleavedContent
+      param_type: ParamType
+      safety_violation: SafetyViolation
+      sampling_params: SamplingParams
+      scoring_result: ScoringResult
+      system_message: SystemMessage
+      query_result: RAGQueryResult
+      document: RAGDocument
+      query_config: RAGQueryConfig
+  toolgroups:
+    models:
+      tool_group: ToolGroup
+      list_tool_groups_response: ListToolGroupsResponse
+    methods:
+      register: post /v1/toolgroups
+      get: get /v1/toolgroups/{toolgroup_id}
+      list: get /v1/toolgroups
+      unregister: delete /v1/toolgroups/{toolgroup_id}
+  tools:
+    methods:
+      get: get /v1/tools/{tool_name}
+      list:
+        endpoint: get /v1/tools
+        paginated: false
+
+  tool_runtime:
+    models:
+      tool_def: ToolDef
+      tool_invocation_result: ToolInvocationResult
+    methods:
+      list_tools:
+        endpoint: get /v1/tool-runtime/list-tools
+        paginated: false
+      invoke_tool: post /v1/tool-runtime/invoke
+    subresources:
+      rag_tool:
+        methods:
+          insert: post /v1/tool-runtime/rag-tool/insert
+          query: post /v1/tool-runtime/rag-tool/query
+
+  responses:
+    models:
+      response_object_stream: OpenAIResponseObjectStream
+      response_object: OpenAIResponseObject
+    methods:
+      create:
+        type: http
+        endpoint: post /v1/responses
+        streaming:
+          stream_event_model: responses.response_object_stream
+          param_discriminator: stream
+      retrieve: get /v1/responses/{response_id}
+      list:
+        type: http
+        endpoint: get /v1/responses
+      delete:
+        type: http
+        endpoint: delete /v1/responses/{response_id}
+    subresources:
+      input_items:
+        methods:
+          list:
+            type: http
+            endpoint: get /v1/responses/{response_id}/input_items
+
+  prompts:
+    models:
+      prompt: Prompt
+      list_prompts_response: ListPromptsResponse
+    methods:
+      create: post /v1/prompts
+      list:
+        endpoint: get /v1/prompts
+        paginated: false
+      retrieve: get /v1/prompts/{prompt_id}
+      update: post /v1/prompts/{prompt_id}
+      delete: delete /v1/prompts/{prompt_id}
+      set_default_version: post /v1/prompts/{prompt_id}/set-default-version
+    subresources:
+      versions:
+        methods:
+          list:
+            endpoint: get /v1/prompts/{prompt_id}/versions
+            paginated: false
+
+  conversations:
+    models:
+      conversation_object: Conversation
+    methods:
+      create:
+        type: http
+        endpoint: post /v1/conversations
+      retrieve: get /v1/conversations/{conversation_id}
+      update:
+        type: http
+        endpoint: post /v1/conversations/{conversation_id}
+      delete:
+        type: http
+        endpoint: delete /v1/conversations/{conversation_id}
+    subresources:
+      items:
+        methods:
+          get:
+            type: http
+            endpoint: get /v1/conversations/{conversation_id}/items/{item_id}
+          list:
+            type: http
+            endpoint: get /v1/conversations/{conversation_id}/items
+          create:
+            type: http
+            endpoint: post /v1/conversations/{conversation_id}/items
+
+  inspect:
+    models:
+      healthInfo: HealthInfo
+      providerInfo: ProviderInfo
+      routeInfo: RouteInfo
+      versionInfo: VersionInfo
+    methods:
+      health: get /v1/health
+      version: get /v1/version
+
+  embeddings:
+    models:
+      create_embeddings_response: OpenAIEmbeddingsResponse
+    methods:
+      create: post /v1/embeddings
+
+  chat:
+    models:
+      chat_completion_chunk: OpenAIChatCompletionChunk
+    subresources:
+      completions:
+        methods:
+          create:
+            type: http
+            endpoint: post /v1/chat/completions
+            streaming:
+              stream_event_model: chat.chat_completion_chunk
+              param_discriminator: stream
+          list:
+            type: http
+            endpoint: get /v1/chat/completions
+          retrieve:
+            type: http
+            endpoint: get /v1/chat/completions/{completion_id}
+  completions:
+    methods:
+      create:
+        type: http
+        endpoint: post /v1/completions
+        streaming:
+          param_discriminator: stream
+
+  vector_io:
+    models:
+      queryChunksResponse: QueryChunksResponse
+    methods:
+      insert: post /v1/vector-io/insert
+      query: post /v1/vector-io/query
+
+  vector_stores:
+    models:
+      vector_store: VectorStoreObject
+      list_vector_stores_response: VectorStoreListResponse
+      vector_store_delete_response: VectorStoreDeleteResponse
+      vector_store_search_response: VectorStoreSearchResponsePage
+    methods:
+      create: post /v1/vector_stores
+      list:
+        endpoint: get /v1/vector_stores
+      retrieve: get /v1/vector_stores/{vector_store_id}
+      update: post /v1/vector_stores/{vector_store_id}
+      delete: delete /v1/vector_stores/{vector_store_id}
+      search: post /v1/vector_stores/{vector_store_id}/search
+    subresources:
+      files:
+        models:
+          vector_store_file: VectorStoreFileObject
+        methods:
+          list: get /v1/vector_stores/{vector_store_id}/files
+          retrieve: get /v1/vector_stores/{vector_store_id}/files/{file_id}
+          update: post /v1/vector_stores/{vector_store_id}/files/{file_id}
+          delete: delete /v1/vector_stores/{vector_store_id}/files/{file_id}
+          create: post /v1/vector_stores/{vector_store_id}/files
+          content: get /v1/vector_stores/{vector_store_id}/files/{file_id}/content
+      file_batches:
+        models:
+          vector_store_file_batches: VectorStoreFileBatchObject
+          list_vector_store_files_in_batch_response: VectorStoreFilesListInBatchResponse
+        methods:
+          create: post /v1/vector_stores/{vector_store_id}/file_batches
+          retrieve: get /v1/vector_stores/{vector_store_id}/file_batches/{batch_id}
+          list_files: get /v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/files
+          cancel: post /v1/vector_stores/{vector_store_id}/file_batches/{batch_id}/cancel
+
+  models:
+    models:
+      model: OpenAIModel
+      list_models_response: OpenAIListModelsResponse
+    methods:
+      list:
+        endpoint: get /v1/models
+        paginated: false
+      retrieve: get /v1/models/{model_id}
+      register: post /v1/models
+      unregister: delete /v1/models/{model_id}
+    subresources:
+      openai:
+        methods:
+          list:
+            endpoint: get /v1/models
+            paginated: false
+
+  providers:
+    models:
+      list_providers_response: ListProvidersResponse
+    methods:
+      list:
+        endpoint: get /v1/providers
+        paginated: false
+      retrieve: get /v1/providers/{provider_id}
+
+  routes:
+    models:
+      list_routes_response: ListRoutesResponse
+    methods:
+      list:
+        endpoint: get /v1/inspect/routes
+        paginated: false
+
+
+  moderations:
+    models:
+      create_response: ModerationObject
+    methods:
+      create: post /v1/moderations
+
+
+  safety:
+    models:
+      run_shield_response: RunShieldResponse
+    methods:
+      run_shield: post /v1/safety/run-shield
+
+
+  shields:
+    models:
+      shield: Shield
+      list_shields_response: ListShieldsResponse
+    methods:
+      retrieve: get /v1/shields/{identifier}
+      list:
+        endpoint: get /v1/shields
+        paginated: false
+      register: post /v1/shields
+      delete: delete /v1/shields/{identifier}
+
+  scoring:
+    methods:
+      score: post /v1/scoring/score
+      score_batch: post /v1/scoring/score-batch
+  scoring_functions:
+    methods:
+      retrieve: get /v1/scoring-functions/{scoring_fn_id}
+      list:
+        endpoint: get /v1/scoring-functions
+        paginated: false
+      register: post /v1/scoring-functions
+    models:
+      scoring_fn: ScoringFn
+      scoring_fn_params: ScoringFnParams
+      list_scoring_functions_response: ListScoringFunctionsResponse
+
+  files:
+    methods:
+      create: post /v1/files
+      list: get /v1/files
+      retrieve: get /v1/files/{file_id}
+      delete: delete /v1/files/{file_id}
+      content: get /v1/files/{file_id}/content
+    models:
+      file: OpenAIFileObject
+      list_files_response: ListOpenAIFileResponse
+      delete_file_response: OpenAIFileDeleteResponse
+
+  alpha:
+    subresources:
+      inference:
+        methods:
+          rerank: post /v1alpha/inference/rerank
+
+      post_training:
+        models:
+          algorithm_config: AlgorithmConfig
+          post_training_job: PostTrainingJob
+          list_post_training_jobs_response: ListPostTrainingJobsResponse
+        methods:
+          preference_optimize: post /v1alpha/post-training/preference-optimize
+          supervised_fine_tune: post /v1alpha/post-training/supervised-fine-tune
+        subresources:
+          job:
+            methods:
+              artifacts: get /v1alpha/post-training/job/artifacts
+              cancel: post /v1alpha/post-training/job/cancel
+              status: get /v1alpha/post-training/job/status
+              list:
+                endpoint: get /v1alpha/post-training/jobs
+                paginated: false
+
+      benchmarks:
+        methods:
+          retrieve: get /v1alpha/eval/benchmarks/{benchmark_id}
+          list:
+            endpoint: get /v1alpha/eval/benchmarks
+            paginated: false
+          register: post /v1alpha/eval/benchmarks
+        models:
+          benchmark: Benchmark
+          list_benchmarks_response: ListBenchmarksResponse
+
+      eval:
+        methods:
+          evaluate_rows: post /v1alpha/eval/benchmarks/{benchmark_id}/evaluations
+          run_eval: post /v1alpha/eval/benchmarks/{benchmark_id}/jobs
+          evaluate_rows_alpha: post /v1alpha/eval/benchmarks/{benchmark_id}/evaluations
+          run_eval_alpha: post /v1alpha/eval/benchmarks/{benchmark_id}/jobs
+
+        subresources:
+          jobs:
+            methods:
+              cancel: delete /v1alpha/eval/benchmarks/{benchmark_id}/jobs/{job_id}
+              status: get /v1alpha/eval/benchmarks/{benchmark_id}/jobs/{job_id}
+              retrieve: get /v1alpha/eval/benchmarks/{benchmark_id}/jobs/{job_id}/result
+        models:
+          evaluate_response: EvaluateResponse
+          benchmark_config: BenchmarkConfig
+          job: Job
+
+  beta:
+    subresources:
+      datasets:
+        models:
+          list_datasets_response: ListDatasetsResponse
+        methods:
+          register: post /v1beta/datasets
+          retrieve: get /v1beta/datasets/{dataset_id}
+          list:
+            endpoint: get /v1beta/datasets
+            paginated: false
+          unregister: delete /v1beta/datasets/{dataset_id}
+          iterrows: get /v1beta/datasetio/iterrows/{dataset_id}
+          appendrows: post /v1beta/datasetio/append-rows/{dataset_id}
+
+
+settings:
+  license: MIT
+  unwrap_response_fields: [ data ]
+
+openapi:
+  transformations:
+    - command: mergeObject
+      reason: Better return_type using enum
+      args:
+        target:
+          - '$.components.schemas'
+        object:
+          ReturnType:
+            additionalProperties: false
+            properties:
+              type:
+                enum:
+                  - string
+                  - number
+                  - boolean
+                  - array
+                  - object
+                  - json
+                  - union
+                  - chat_completion_input
+                  - completion_input
+                  - agent_turn_input
+            required:
+              - type
+            type: object
+    - command: replaceProperties
+      reason: Replace return type properties with better model (see above)
+      args:
+        filter:
+          only:
+            - '$.components.schemas.ScoringFn.properties.return_type'
+            - '$.components.schemas.RegisterScoringFunctionRequest.properties.return_type'
+        value:
+          $ref: '#/components/schemas/ReturnType'
+    - command: oneOfToAnyOf
+      reason: Prism (mock server) doesn't like one of our requests as it technically matches multiple variants
+
+# `readme` is used to configure the code snippets that will be rendered in the
+# README.md of various SDKs. In particular, you can change the `headline`
+# snippet's endpoint and the arguments to call it with.
+readme:
+  example_requests:
+    default:
+      type: request
+      endpoint: post /v1/chat/completions
+      params: &ref_0 {}
+    headline:
+      type: request
+      endpoint: post /v1/models
+      params: *ref_0
+    pagination:
+      type: request
+      endpoint: post /v1/chat/completions
+      params: {}

client-sdks/stainless/openapi.yml

@@ -2055,69 +2055,6 @@ paths:
           schema:
             $ref: '#/components/schemas/URL'
       deprecated: false
-  /v1/tool-runtime/rag-tool/insert:
-    post:
-      responses:
-        '200':
-          description: OK
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Index documents so they can be used by the RAG system.
-      description: >-
-        Index documents so they can be used by the RAG system.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/InsertRequest'
-        required: true
-      deprecated: false
-  /v1/tool-runtime/rag-tool/query:
-    post:
-      responses:
-        '200':
-          description: >-
-            RAGQueryResult containing the retrieved content and metadata
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/RAGQueryResult'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Query the RAG system for context; typically invoked by the agent.
-      description: >-
-        Query the RAG system for context; typically invoked by the agent.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/QueryRequest'
-        required: true
-      deprecated: false
   /v1/toolgroups:
     get:
       responses:
@@ -9633,274 +9570,6 @@ components:
       title: ListToolDefsResponse
       description: >-
         Response containing a list of tool definitions.
-    RAGDocument:
-      type: object
-      properties:
-        document_id:
-          type: string
-          description: The unique identifier for the document.
-        content:
-          oneOf:
-            - type: string
-            - $ref: '#/components/schemas/InterleavedContentItem'
-            - type: array
-              items:
-                $ref: '#/components/schemas/InterleavedContentItem'
-            - $ref: '#/components/schemas/URL'
-          description: The content of the document.
-        mime_type:
-          type: string
-          description: The MIME type of the document.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Additional metadata for the document.
-      additionalProperties: false
-      required:
-        - document_id
-        - content
-        - metadata
-      title: RAGDocument
-      description: >-
-        A document to be used for document ingestion in the RAG Tool.
-    InsertRequest:
-      type: object
-      properties:
-        documents:
-          type: array
-          items:
-            $ref: '#/components/schemas/RAGDocument'
-          description: >-
-            List of documents to index in the RAG system
-        vector_store_id:
-          type: string
-          description: >-
-            ID of the vector database to store the document embeddings
-        chunk_size_in_tokens:
-          type: integer
-          description: >-
-            (Optional) Size in tokens for document chunking during indexing
-      additionalProperties: false
-      required:
-        - documents
-        - vector_store_id
-        - chunk_size_in_tokens
-      title: InsertRequest
-    DefaultRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: default
-          default: default
-          description: >-
-            Type of query generator, always 'default'
-        separator:
-          type: string
-          default: ' '
-          description: >-
-            String separator used to join query terms
-      additionalProperties: false
-      required:
-        - type
-        - separator
-      title: DefaultRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the default RAG query generator.
-    LLMRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: llm
-          default: llm
-          description: Type of query generator, always 'llm'
-        model:
-          type: string
-          description: >-
-            Name of the language model to use for query generation
-        template:
-          type: string
-          description: >-
-            Template string for formatting the query generation prompt
-      additionalProperties: false
-      required:
-        - type
-        - model
-        - template
-      title: LLMRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the LLM-based RAG query generator.
-    RAGQueryConfig:
-      type: object
-      properties:
-        query_generator_config:
-          oneOf:
-            - $ref: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-            - $ref: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          discriminator:
-            propertyName: type
-            mapping:
-              default: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-              llm: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          description: Configuration for the query generator.
-        max_tokens_in_context:
-          type: integer
-          default: 4096
-          description: Maximum number of tokens in the context.
-        max_chunks:
-          type: integer
-          default: 5
-          description: Maximum number of chunks to retrieve.
-        chunk_template:
-          type: string
-          default: >
-            Result {index}
-
-            Content: {chunk.content}
-
-            Metadata: {metadata}
-          description: >-
-            Template for formatting each retrieved chunk in the context. Available
-            placeholders: {index} (1-based chunk ordinal), {chunk.content} (chunk
-            content string), {metadata} (chunk metadata dict). Default: "Result {index}\nContent:
-            {chunk.content}\nMetadata: {metadata}\n"
-        mode:
-          $ref: '#/components/schemas/RAGSearchMode'
-          default: vector
-          description: >-
-            Search mode for retrieval—either "vector", "keyword", or "hybrid". Default
-            "vector".
-        ranker:
-          $ref: '#/components/schemas/Ranker'
-          description: >-
-            Configuration for the ranker to use in hybrid search. Defaults to RRF
-            ranker.
-      additionalProperties: false
-      required:
-        - query_generator_config
-        - max_tokens_in_context
-        - max_chunks
-        - chunk_template
-      title: RAGQueryConfig
-      description: >-
-        Configuration for the RAG query generation.
-    RAGSearchMode:
-      type: string
-      enum:
-        - vector
-        - keyword
-        - hybrid
-      title: RAGSearchMode
-      description: >-
-        Search modes for RAG query retrieval: - VECTOR: Uses vector similarity search
-        for semantic matching - KEYWORD: Uses keyword-based search for exact matching
-        - HYBRID: Combines both vector and keyword search for better results
-    RRFRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: rrf
-          default: rrf
-          description: The type of ranker, always "rrf"
-        impact_factor:
-          type: number
-          default: 60.0
-          description: >-
-            The impact factor for RRF scoring. Higher values give more weight to higher-ranked
-            results. Must be greater than 0
-      additionalProperties: false
-      required:
-        - type
-        - impact_factor
-      title: RRFRanker
-      description: >-
-        Reciprocal Rank Fusion (RRF) ranker configuration.
-    Ranker:
-      oneOf:
-        - $ref: '#/components/schemas/RRFRanker'
-        - $ref: '#/components/schemas/WeightedRanker'
-      discriminator:
-        propertyName: type
-        mapping:
-          rrf: '#/components/schemas/RRFRanker'
-          weighted: '#/components/schemas/WeightedRanker'
-    WeightedRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: weighted
-          default: weighted
-          description: The type of ranker, always "weighted"
-        alpha:
-          type: number
-          default: 0.5
-          description: >-
-            Weight factor between 0 and 1. 0 means only use keyword scores, 1 means
-            only use vector scores, values in between blend both scores.
-      additionalProperties: false
-      required:
-        - type
-        - alpha
-      title: WeightedRanker
-      description: >-
-        Weighted ranker configuration that combines vector and keyword scores.
-    QueryRequest:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The query content to search for in the indexed documents
-        vector_store_ids:
-          type: array
-          items:
-            type: string
-          description: >-
-            List of vector database IDs to search within
-        query_config:
-          $ref: '#/components/schemas/RAGQueryConfig'
-          description: >-
-            (Optional) Configuration parameters for the query operation
-      additionalProperties: false
-      required:
-        - content
-        - vector_store_ids
-      title: QueryRequest
-    RAGQueryResult:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) The retrieved content from the query
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            Additional metadata about the query result
-      additionalProperties: false
-      required:
-        - metadata
-      title: RAGQueryResult
-      description: >-
-        Result of a RAG query containing retrieved content and metadata.
     ToolGroup:
       type: object
       properties:

docs/docs/distributions/k8s/ui-k8s.yaml.template

@@ -44,7 +44,7 @@ spec:
 
             # Navigate to the UI directory
             echo "Navigating to UI directory..."
-            cd /app/llama_stack/ui
+            cd /app/llama_stack_ui
 
             # Check if package.json exists
             if [ ! -f "package.json" ]; then

docs/openapi_generator/pyopenapi/operations.py

@@ -196,16 +196,10 @@ def _get_endpoint_functions(
 def _get_defining_class(member_fn: str, derived_cls: type) -> type:
     "Find the class in which a member function is first defined in a class inheritance hierarchy."
 
-    # This import must be dynamic here
-    from llama_stack.apis.tools import RAGToolRuntime, ToolRuntime
-
     # iterate in reverse member resolution order to find most specific class first
     for cls in reversed(inspect.getmro(derived_cls)):
         for name, _ in inspect.getmembers(cls, inspect.isfunction):
             if name == member_fn:
-                # HACK ALERT
-                if cls == RAGToolRuntime:
-                    return ToolRuntime
                 return cls
 
     raise ValidationError(

docs/static/llama-stack-spec.yaml

@@ -2052,69 +2052,6 @@ paths:
           schema:
             $ref: '#/components/schemas/URL'
       deprecated: false
-  /v1/tool-runtime/rag-tool/insert:
-    post:
-      responses:
-        '200':
-          description: OK
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Index documents so they can be used by the RAG system.
-      description: >-
-        Index documents so they can be used by the RAG system.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/InsertRequest'
-        required: true
-      deprecated: false
-  /v1/tool-runtime/rag-tool/query:
-    post:
-      responses:
-        '200':
-          description: >-
-            RAGQueryResult containing the retrieved content and metadata
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/RAGQueryResult'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Query the RAG system for context; typically invoked by the agent.
-      description: >-
-        Query the RAG system for context; typically invoked by the agent.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/QueryRequest'
-        required: true
-      deprecated: false
   /v1/toolgroups:
     get:
       responses:
@@ -8917,274 +8854,6 @@ components:
       title: ListToolDefsResponse
       description: >-
         Response containing a list of tool definitions.
-    RAGDocument:
-      type: object
-      properties:
-        document_id:
-          type: string
-          description: The unique identifier for the document.
-        content:
-          oneOf:
-            - type: string
-            - $ref: '#/components/schemas/InterleavedContentItem'
-            - type: array
-              items:
-                $ref: '#/components/schemas/InterleavedContentItem'
-            - $ref: '#/components/schemas/URL'
-          description: The content of the document.
-        mime_type:
-          type: string
-          description: The MIME type of the document.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Additional metadata for the document.
-      additionalProperties: false
-      required:
-        - document_id
-        - content
-        - metadata
-      title: RAGDocument
-      description: >-
-        A document to be used for document ingestion in the RAG Tool.
-    InsertRequest:
-      type: object
-      properties:
-        documents:
-          type: array
-          items:
-            $ref: '#/components/schemas/RAGDocument'
-          description: >-
-            List of documents to index in the RAG system
-        vector_store_id:
-          type: string
-          description: >-
-            ID of the vector database to store the document embeddings
-        chunk_size_in_tokens:
-          type: integer
-          description: >-
-            (Optional) Size in tokens for document chunking during indexing
-      additionalProperties: false
-      required:
-        - documents
-        - vector_store_id
-        - chunk_size_in_tokens
-      title: InsertRequest
-    DefaultRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: default
-          default: default
-          description: >-
-            Type of query generator, always 'default'
-        separator:
-          type: string
-          default: ' '
-          description: >-
-            String separator used to join query terms
-      additionalProperties: false
-      required:
-        - type
-        - separator
-      title: DefaultRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the default RAG query generator.
-    LLMRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: llm
-          default: llm
-          description: Type of query generator, always 'llm'
-        model:
-          type: string
-          description: >-
-            Name of the language model to use for query generation
-        template:
-          type: string
-          description: >-
-            Template string for formatting the query generation prompt
-      additionalProperties: false
-      required:
-        - type
-        - model
-        - template
-      title: LLMRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the LLM-based RAG query generator.
-    RAGQueryConfig:
-      type: object
-      properties:
-        query_generator_config:
-          oneOf:
-            - $ref: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-            - $ref: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          discriminator:
-            propertyName: type
-            mapping:
-              default: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-              llm: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          description: Configuration for the query generator.
-        max_tokens_in_context:
-          type: integer
-          default: 4096
-          description: Maximum number of tokens in the context.
-        max_chunks:
-          type: integer
-          default: 5
-          description: Maximum number of chunks to retrieve.
-        chunk_template:
-          type: string
-          default: >
-            Result {index}
-
-            Content: {chunk.content}
-
-            Metadata: {metadata}
-          description: >-
-            Template for formatting each retrieved chunk in the context. Available
-            placeholders: {index} (1-based chunk ordinal), {chunk.content} (chunk
-            content string), {metadata} (chunk metadata dict). Default: "Result {index}\nContent:
-            {chunk.content}\nMetadata: {metadata}\n"
-        mode:
-          $ref: '#/components/schemas/RAGSearchMode'
-          default: vector
-          description: >-
-            Search mode for retrieval—either "vector", "keyword", or "hybrid". Default
-            "vector".
-        ranker:
-          $ref: '#/components/schemas/Ranker'
-          description: >-
-            Configuration for the ranker to use in hybrid search. Defaults to RRF
-            ranker.
-      additionalProperties: false
-      required:
-        - query_generator_config
-        - max_tokens_in_context
-        - max_chunks
-        - chunk_template
-      title: RAGQueryConfig
-      description: >-
-        Configuration for the RAG query generation.
-    RAGSearchMode:
-      type: string
-      enum:
-        - vector
-        - keyword
-        - hybrid
-      title: RAGSearchMode
-      description: >-
-        Search modes for RAG query retrieval: - VECTOR: Uses vector similarity search
-        for semantic matching - KEYWORD: Uses keyword-based search for exact matching
-        - HYBRID: Combines both vector and keyword search for better results
-    RRFRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: rrf
-          default: rrf
-          description: The type of ranker, always "rrf"
-        impact_factor:
-          type: number
-          default: 60.0
-          description: >-
-            The impact factor for RRF scoring. Higher values give more weight to higher-ranked
-            results. Must be greater than 0
-      additionalProperties: false
-      required:
-        - type
-        - impact_factor
-      title: RRFRanker
-      description: >-
-        Reciprocal Rank Fusion (RRF) ranker configuration.
-    Ranker:
-      oneOf:
-        - $ref: '#/components/schemas/RRFRanker'
-        - $ref: '#/components/schemas/WeightedRanker'
-      discriminator:
-        propertyName: type
-        mapping:
-          rrf: '#/components/schemas/RRFRanker'
-          weighted: '#/components/schemas/WeightedRanker'
-    WeightedRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: weighted
-          default: weighted
-          description: The type of ranker, always "weighted"
-        alpha:
-          type: number
-          default: 0.5
-          description: >-
-            Weight factor between 0 and 1. 0 means only use keyword scores, 1 means
-            only use vector scores, values in between blend both scores.
-      additionalProperties: false
-      required:
-        - type
-        - alpha
-      title: WeightedRanker
-      description: >-
-        Weighted ranker configuration that combines vector and keyword scores.
-    QueryRequest:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The query content to search for in the indexed documents
-        vector_store_ids:
-          type: array
-          items:
-            type: string
-          description: >-
-            List of vector database IDs to search within
-        query_config:
-          $ref: '#/components/schemas/RAGQueryConfig'
-          description: >-
-            (Optional) Configuration parameters for the query operation
-      additionalProperties: false
-      required:
-        - content
-        - vector_store_ids
-      title: QueryRequest
-    RAGQueryResult:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) The retrieved content from the query
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            Additional metadata about the query result
-      additionalProperties: false
-      required:
-        - metadata
-      title: RAGQueryResult
-      description: >-
-        Result of a RAG query containing retrieved content and metadata.
     ToolGroup:
       type: object
       properties:

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

@@ -2055,69 +2055,6 @@ paths:
           schema:
             $ref: '#/components/schemas/URL'
       deprecated: false
-  /v1/tool-runtime/rag-tool/insert:
-    post:
-      responses:
-        '200':
-          description: OK
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Index documents so they can be used by the RAG system.
-      description: >-
-        Index documents so they can be used by the RAG system.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/InsertRequest'
-        required: true
-      deprecated: false
-  /v1/tool-runtime/rag-tool/query:
-    post:
-      responses:
-        '200':
-          description: >-
-            RAGQueryResult containing the retrieved content and metadata
-          content:
-            application/json:
-              schema:
-                $ref: '#/components/schemas/RAGQueryResult'
-        '400':
-          $ref: '#/components/responses/BadRequest400'
-        '429':
-          $ref: >-
-            #/components/responses/TooManyRequests429
-        '500':
-          $ref: >-
-            #/components/responses/InternalServerError500
-        default:
-          $ref: '#/components/responses/DefaultError'
-      tags:
-        - ToolRuntime
-      summary: >-
-        Query the RAG system for context; typically invoked by the agent.
-      description: >-
-        Query the RAG system for context; typically invoked by the agent.
-      parameters: []
-      requestBody:
-        content:
-          application/json:
-            schema:
-              $ref: '#/components/schemas/QueryRequest'
-        required: true
-      deprecated: false
   /v1/toolgroups:
     get:
       responses:
@@ -9633,274 +9570,6 @@ components:
       title: ListToolDefsResponse
       description: >-
         Response containing a list of tool definitions.
-    RAGDocument:
-      type: object
-      properties:
-        document_id:
-          type: string
-          description: The unique identifier for the document.
-        content:
-          oneOf:
-            - type: string
-            - $ref: '#/components/schemas/InterleavedContentItem'
-            - type: array
-              items:
-                $ref: '#/components/schemas/InterleavedContentItem'
-            - $ref: '#/components/schemas/URL'
-          description: The content of the document.
-        mime_type:
-          type: string
-          description: The MIME type of the document.
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: Additional metadata for the document.
-      additionalProperties: false
-      required:
-        - document_id
-        - content
-        - metadata
-      title: RAGDocument
-      description: >-
-        A document to be used for document ingestion in the RAG Tool.
-    InsertRequest:
-      type: object
-      properties:
-        documents:
-          type: array
-          items:
-            $ref: '#/components/schemas/RAGDocument'
-          description: >-
-            List of documents to index in the RAG system
-        vector_store_id:
-          type: string
-          description: >-
-            ID of the vector database to store the document embeddings
-        chunk_size_in_tokens:
-          type: integer
-          description: >-
-            (Optional) Size in tokens for document chunking during indexing
-      additionalProperties: false
-      required:
-        - documents
-        - vector_store_id
-        - chunk_size_in_tokens
-      title: InsertRequest
-    DefaultRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: default
-          default: default
-          description: >-
-            Type of query generator, always 'default'
-        separator:
-          type: string
-          default: ' '
-          description: >-
-            String separator used to join query terms
-      additionalProperties: false
-      required:
-        - type
-        - separator
-      title: DefaultRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the default RAG query generator.
-    LLMRAGQueryGeneratorConfig:
-      type: object
-      properties:
-        type:
-          type: string
-          const: llm
-          default: llm
-          description: Type of query generator, always 'llm'
-        model:
-          type: string
-          description: >-
-            Name of the language model to use for query generation
-        template:
-          type: string
-          description: >-
-            Template string for formatting the query generation prompt
-      additionalProperties: false
-      required:
-        - type
-        - model
-        - template
-      title: LLMRAGQueryGeneratorConfig
-      description: >-
-        Configuration for the LLM-based RAG query generator.
-    RAGQueryConfig:
-      type: object
-      properties:
-        query_generator_config:
-          oneOf:
-            - $ref: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-            - $ref: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          discriminator:
-            propertyName: type
-            mapping:
-              default: '#/components/schemas/DefaultRAGQueryGeneratorConfig'
-              llm: '#/components/schemas/LLMRAGQueryGeneratorConfig'
-          description: Configuration for the query generator.
-        max_tokens_in_context:
-          type: integer
-          default: 4096
-          description: Maximum number of tokens in the context.
-        max_chunks:
-          type: integer
-          default: 5
-          description: Maximum number of chunks to retrieve.
-        chunk_template:
-          type: string
-          default: >
-            Result {index}
-
-            Content: {chunk.content}
-
-            Metadata: {metadata}
-          description: >-
-            Template for formatting each retrieved chunk in the context. Available
-            placeholders: {index} (1-based chunk ordinal), {chunk.content} (chunk
-            content string), {metadata} (chunk metadata dict). Default: "Result {index}\nContent:
-            {chunk.content}\nMetadata: {metadata}\n"
-        mode:
-          $ref: '#/components/schemas/RAGSearchMode'
-          default: vector
-          description: >-
-            Search mode for retrieval—either "vector", "keyword", or "hybrid". Default
-            "vector".
-        ranker:
-          $ref: '#/components/schemas/Ranker'
-          description: >-
-            Configuration for the ranker to use in hybrid search. Defaults to RRF
-            ranker.
-      additionalProperties: false
-      required:
-        - query_generator_config
-        - max_tokens_in_context
-        - max_chunks
-        - chunk_template
-      title: RAGQueryConfig
-      description: >-
-        Configuration for the RAG query generation.
-    RAGSearchMode:
-      type: string
-      enum:
-        - vector
-        - keyword
-        - hybrid
-      title: RAGSearchMode
-      description: >-
-        Search modes for RAG query retrieval: - VECTOR: Uses vector similarity search
-        for semantic matching - KEYWORD: Uses keyword-based search for exact matching
-        - HYBRID: Combines both vector and keyword search for better results
-    RRFRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: rrf
-          default: rrf
-          description: The type of ranker, always "rrf"
-        impact_factor:
-          type: number
-          default: 60.0
-          description: >-
-            The impact factor for RRF scoring. Higher values give more weight to higher-ranked
-            results. Must be greater than 0
-      additionalProperties: false
-      required:
-        - type
-        - impact_factor
-      title: RRFRanker
-      description: >-
-        Reciprocal Rank Fusion (RRF) ranker configuration.
-    Ranker:
-      oneOf:
-        - $ref: '#/components/schemas/RRFRanker'
-        - $ref: '#/components/schemas/WeightedRanker'
-      discriminator:
-        propertyName: type
-        mapping:
-          rrf: '#/components/schemas/RRFRanker'
-          weighted: '#/components/schemas/WeightedRanker'
-    WeightedRanker:
-      type: object
-      properties:
-        type:
-          type: string
-          const: weighted
-          default: weighted
-          description: The type of ranker, always "weighted"
-        alpha:
-          type: number
-          default: 0.5
-          description: >-
-            Weight factor between 0 and 1. 0 means only use keyword scores, 1 means
-            only use vector scores, values in between blend both scores.
-      additionalProperties: false
-      required:
-        - type
-        - alpha
-      title: WeightedRanker
-      description: >-
-        Weighted ranker configuration that combines vector and keyword scores.
-    QueryRequest:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            The query content to search for in the indexed documents
-        vector_store_ids:
-          type: array
-          items:
-            type: string
-          description: >-
-            List of vector database IDs to search within
-        query_config:
-          $ref: '#/components/schemas/RAGQueryConfig'
-          description: >-
-            (Optional) Configuration parameters for the query operation
-      additionalProperties: false
-      required:
-        - content
-        - vector_store_ids
-      title: QueryRequest
-    RAGQueryResult:
-      type: object
-      properties:
-        content:
-          $ref: '#/components/schemas/InterleavedContent'
-          description: >-
-            (Optional) The retrieved content from the query
-        metadata:
-          type: object
-          additionalProperties:
-            oneOf:
-              - type: 'null'
-              - type: boolean
-              - type: number
-              - type: string
-              - type: array
-              - type: object
-          description: >-
-            Additional metadata about the query result
-      additionalProperties: false
-      required:
-        - metadata
-      title: RAGQueryResult
-      description: >-
-        Result of a RAG query containing retrieved content and metadata.
     ToolGroup:
       type: object
       properties:

scripts/run-ui-linter.sh

@@ -6,7 +6,7 @@
 # the root directory of this source tree.
 
 set -e
-cd src/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"

src/llama_stack/apis/tools/rag_tool.py

@@ -5,18 +5,13 @@
 # the root directory of this source tree.
 
 from enum import Enum, StrEnum
-from typing import Annotated, Any, Literal, Protocol
+from typing import Annotated, Any, Literal
 
 from pydantic import BaseModel, Field, field_validator
-from typing_extensions import runtime_checkable
 
 from llama_stack.apis.common.content_types import URL, InterleavedContent
-from llama_stack.apis.common.tracing import trace_protocol
-from llama_stack.apis.version import LLAMA_STACK_API_V1
-from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
 
 
-@json_schema_type
 class RRFRanker(BaseModel):
     """
     Reciprocal Rank Fusion (RRF) ranker configuration.
@@ -30,7 +25,6 @@ class RRFRanker(BaseModel):
     impact_factor: float = Field(default=60.0, gt=0.0)  # default of 60 for optimal performance
 
 
-@json_schema_type
 class WeightedRanker(BaseModel):
     """
     Weighted ranker configuration that combines vector and keyword scores.
@@ -55,10 +49,8 @@ Ranker = Annotated[
     RRFRanker | WeightedRanker,
     Field(discriminator="type"),
 ]
-register_schema(Ranker, name="Ranker")
 
 
-@json_schema_type
 class RAGDocument(BaseModel):
     """
     A document to be used for document ingestion in the RAG Tool.
@@ -75,7 +67,6 @@ class RAGDocument(BaseModel):
     metadata: dict[str, Any] = Field(default_factory=dict)
 
 
-@json_schema_type
 class RAGQueryResult(BaseModel):
     """Result of a RAG query containing retrieved content and metadata.
 
@@ -87,7 +78,6 @@ class RAGQueryResult(BaseModel):
     metadata: dict[str, Any] = Field(default_factory=dict)
 
 
-@json_schema_type
 class RAGQueryGenerator(Enum):
     """Types of query generators for RAG systems.
 
@@ -101,7 +91,6 @@ class RAGQueryGenerator(Enum):
     custom = "custom"
 
 
-@json_schema_type
 class RAGSearchMode(StrEnum):
     """
     Search modes for RAG query retrieval:
@@ -115,7 +104,6 @@ class RAGSearchMode(StrEnum):
     HYBRID = "hybrid"
 
 
-@json_schema_type
 class DefaultRAGQueryGeneratorConfig(BaseModel):
     """Configuration for the default RAG query generator.
 
@@ -127,7 +115,6 @@ class DefaultRAGQueryGeneratorConfig(BaseModel):
     separator: str = " "
 
 
-@json_schema_type
 class LLMRAGQueryGeneratorConfig(BaseModel):
     """Configuration for the LLM-based RAG query generator.
 
@@ -145,10 +132,8 @@ RAGQueryGeneratorConfig = Annotated[
     DefaultRAGQueryGeneratorConfig | LLMRAGQueryGeneratorConfig,
     Field(discriminator="type"),
 ]
-register_schema(RAGQueryGeneratorConfig, name="RAGQueryGeneratorConfig")
 
 
-@json_schema_type
 class RAGQueryConfig(BaseModel):
     """
     Configuration for the RAG query generation.
@@ -181,38 +166,3 @@ class RAGQueryConfig(BaseModel):
         if len(v) == 0:
             raise ValueError("chunk_template must not be empty")
         return v
-
-
-@runtime_checkable
-@trace_protocol
-class RAGToolRuntime(Protocol):
-    @webmethod(route="/tool-runtime/rag-tool/insert", method="POST", level=LLAMA_STACK_API_V1)
-    async def insert(
-        self,
-        documents: list[RAGDocument],
-        vector_store_id: str,
-        chunk_size_in_tokens: int = 512,
-    ) -> None:
-        """Index documents so they can be used by the RAG system.
-
-        :param documents: List of documents to index in the RAG system
-        :param vector_store_id: ID of the vector database to store the document embeddings
-        :param chunk_size_in_tokens: (Optional) Size in tokens for document chunking during indexing
-        """
-        ...
-
-    @webmethod(route="/tool-runtime/rag-tool/query", method="POST", level=LLAMA_STACK_API_V1)
-    async def query(
-        self,
-        content: InterleavedContent,
-        vector_store_ids: list[str],
-        query_config: RAGQueryConfig | None = None,
-    ) -> RAGQueryResult:
-        """Query the RAG system for context; typically invoked by the agent.
-
-        :param content: The query content to search for in the indexed documents
-        :param vector_store_ids: List of vector database IDs to search within
-        :param query_config: (Optional) Configuration parameters for the query operation
-        :returns: RAGQueryResult containing the retrieved content and metadata
-        """
-        ...

src/llama_stack/apis/tools/tools.py

@@ -16,8 +16,6 @@ from llama_stack.apis.resource import Resource, ResourceType
 from llama_stack.apis.version import LLAMA_STACK_API_V1
 from llama_stack.schema_utils import json_schema_type, webmethod
 
-from .rag_tool import RAGToolRuntime
-
 
 @json_schema_type
 class ToolDef(BaseModel):
@@ -195,8 +193,6 @@ class SpecialToolGroup(Enum):
 class ToolRuntime(Protocol):
     tool_store: ToolStore | None = None
 
-    rag_tool: RAGToolRuntime | None = None
-
     # TODO: This needs to be renamed once OPEN API generator name conflict issue is fixed.
     @webmethod(route="/tool-runtime/list-tools", method="GET", level=LLAMA_STACK_API_V1)
     async def list_runtime_tools(

src/llama_stack/cli/stack/run.py

@@ -253,7 +253,7 @@ class StackRun(Subcommand):
             )
             return
 
-        ui_dir = REPO_ROOT / "llama_stack" / "ui"
+        ui_dir = REPO_ROOT / "llama_stack_ui"
         logs_dir = Path("~/.llama/ui/logs").expanduser()
         try:
             # Create logs directory if it doesn't exist

src/llama_stack/core/routers/tool_runtime.py

@@ -8,14 +8,9 @@ from typing import Any
 
 from llama_stack.apis.common.content_types import (
     URL,
-    InterleavedContent,
 )
 from llama_stack.apis.tools import (
     ListToolDefsResponse,
-    RAGDocument,
-    RAGQueryConfig,
-    RAGQueryResult,
-    RAGToolRuntime,
     ToolRuntime,
 )
 from llama_stack.log import get_logger
@@ -26,36 +21,6 @@ logger = get_logger(name=__name__, category="core::routers")
 
 
 class ToolRuntimeRouter(ToolRuntime):
-    class RagToolImpl(RAGToolRuntime):
-        def __init__(
-            self,
-            routing_table: ToolGroupsRoutingTable,
-        ) -> None:
-            logger.debug("Initializing ToolRuntimeRouter.RagToolImpl")
-            self.routing_table = routing_table
-
-        async def query(
-            self,
-            content: InterleavedContent,
-            vector_store_ids: list[str],
-            query_config: RAGQueryConfig | None = None,
-        ) -> RAGQueryResult:
-            logger.debug(f"ToolRuntimeRouter.RagToolImpl.query: {vector_store_ids}")
-            provider = await self.routing_table.get_provider_impl("knowledge_search")
-            return await provider.query(content, vector_store_ids, query_config)
-
-        async def insert(
-            self,
-            documents: list[RAGDocument],
-            vector_store_id: str,
-            chunk_size_in_tokens: int = 512,
-        ) -> None:
-            logger.debug(
-                f"ToolRuntimeRouter.RagToolImpl.insert: {vector_store_id}, {len(documents)} documents, chunk_size={chunk_size_in_tokens}"
-            )
-            provider = await self.routing_table.get_provider_impl("insert_into_memory")
-            return await provider.insert(documents, vector_store_id, chunk_size_in_tokens)
-
     def __init__(
         self,
         routing_table: ToolGroupsRoutingTable,
@@ -63,11 +28,6 @@ class ToolRuntimeRouter(ToolRuntime):
         logger.debug("Initializing ToolRuntimeRouter")
         self.routing_table = routing_table
 
-        # HACK ALERT this should be in sync with "get_all_api_endpoints()"
-        self.rag_tool = self.RagToolImpl(routing_table)
-        for method in ("query", "insert"):
-            setattr(self, f"rag_tool.{method}", getattr(self.rag_tool, method))
-
     async def initialize(self) -> None:
         logger.debug("ToolRuntimeRouter.initialize")
         pass

src/llama_stack/core/server/routes.py

@@ -13,7 +13,6 @@ from aiohttp import hdrs
 from starlette.routing import Route
 
 from llama_stack.apis.datatypes import Api, ExternalApiSpec
-from llama_stack.apis.tools import RAGToolRuntime, SpecialToolGroup
 from llama_stack.core.resolver import api_protocol_map
 from llama_stack.schema_utils import WebMethod
 
@@ -25,33 +24,16 @@ RouteImpls = dict[str, PathImpl]
 RouteMatch = tuple[EndpointFunc, PathParams, str, WebMethod]
 
 
-def toolgroup_protocol_map():
-    return {
-        SpecialToolGroup.rag_tool: RAGToolRuntime,
-    }
-
-
 def get_all_api_routes(
     external_apis: dict[Api, ExternalApiSpec] | None = None,
 ) -> dict[Api, list[tuple[Route, WebMethod]]]:
     apis = {}
 
     protocols = api_protocol_map(external_apis)
-    toolgroup_protocols = toolgroup_protocol_map()
     for api, protocol in protocols.items():
         routes = []
         protocol_methods = inspect.getmembers(protocol, predicate=inspect.isfunction)
 
-        # HACK ALERT
-        if api == Api.tool_runtime:
-            for tool_group in SpecialToolGroup:
-                sub_protocol = toolgroup_protocols[tool_group]
-                sub_protocol_methods = inspect.getmembers(sub_protocol, predicate=inspect.isfunction)
-                for name, method in sub_protocol_methods:
-                    if not hasattr(method, "__webmethod__"):
-                        continue
-                    protocol_methods.append((f"{tool_group.value}.{name}", method))
-
         for name, method in protocol_methods:
             # Get all webmethods for this method (supports multiple decorators)
             webmethods = getattr(method, "__webmethods__", [])

src/llama_stack/core/stack.py

@@ -31,7 +31,7 @@ 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.tools import RAGToolRuntime, ToolGroups, ToolRuntime
+from llama_stack.apis.tools import ToolGroups, ToolRuntime
 from llama_stack.apis.vector_io import VectorIO
 from llama_stack.core.conversations.conversations import ConversationServiceConfig, ConversationServiceImpl
 from llama_stack.core.datatypes import Provider, SafetyConfig, StackRunConfig, VectorStoresConfig
@@ -78,7 +78,6 @@ class LlamaStack(
     Inspect,
     ToolGroups,
     ToolRuntime,
-    RAGToolRuntime,
     Files,
     Prompts,
     Conversations,

src/llama_stack/providers/inline/tool_runtime/rag/memory.py

@@ -27,7 +27,6 @@ from llama_stack.apis.tools import (
     RAGDocument,
     RAGQueryConfig,
     RAGQueryResult,
-    RAGToolRuntime,
     ToolDef,
     ToolGroup,
     ToolInvocationResult,
@@ -91,7 +90,7 @@ async def raw_data_from_doc(doc: RAGDocument) -> tuple[bytes, str]:
             return content_str.encode("utf-8"), "text/plain"
 
 
-class MemoryToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime, RAGToolRuntime):
+class MemoryToolRuntimeImpl(ToolGroupsProtocolPrivate, ToolRuntime):
     def __init__(
         self,
         config: RagToolRuntimeConfig,