feat: split API and provider specs into separate llama-stack-api pkg (#3895)

# What does this PR do?

Extract API definitions and provider specifications into a standalone
llama-stack-api package that can be published to PyPI independently of
the main llama-stack server.


see: https://github.com/llamastack/llama-stack/pull/2978 and
https://github.com/llamastack/llama-stack/pull/2978#issuecomment-3145115942

Motivation

External providers currently import from llama-stack, which overrides
the installed version and causes dependency conflicts. This separation
allows external providers to:

- Install only the type definitions they need without server
dependencies
- Avoid version conflicts with the installed llama-stack package
- Be versioned and released independently

This enables us to re-enable external provider module tests that were
previously blocked by these import conflicts.

Changes

- Created llama-stack-api package with minimal dependencies (pydantic,
jsonschema)
- Moved APIs, providers datatypes, strong_typing, and schema_utils
- Updated all imports from llama_stack.* to llama_stack_api.*
- Configured local editable install for development workflow
- Updated linting and type-checking configuration for both packages

Next Steps

- Publish llama-stack-api to PyPI
- Update external provider dependencies
- Re-enable external provider module tests


Pre-cursor PRs to this one:

- #4093 
- #3954 
- #4064 

These PRs moved key pieces _out_ of the Api pkg, limiting the scope of
change here.


relates to #3237 

## Test Plan

Package builds successfully and can be imported independently. All
pre-commit hooks pass with expected exclusions maintained.

---------

Signed-off-by: Charlie Doern <cdoern@redhat.com>

src/llama_stack/core/build.py

@@ -6,6 +6,7 @@
 
 import sys
 
+from llama_stack_api import Api
 from pydantic import BaseModel
 from termcolor import cprint
 
@@ -13,7 +14,6 @@ from llama_stack.core.datatypes import BuildConfig
 from llama_stack.core.distribution import get_provider_registry
 from llama_stack.distributions.template import DistributionTemplate
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import Api
 
 log = get_logger(name=__name__, category="core")
 

src/llama_stack/core/client.py

@@ -12,11 +12,10 @@ from enum import Enum
 from typing import Any, Union, get_args, get_origin
 
 import httpx
+from llama_stack_api import RemoteProviderConfig
 from pydantic import BaseModel, parse_obj_as
 from termcolor import cprint
 
-from llama_stack.providers.datatypes import RemoteProviderConfig
-
 _CLIENT_CLASSES = {}
 
 

src/llama_stack/core/configure.py

@@ -6,6 +6,8 @@
 import textwrap
 from typing import Any
 
+from llama_stack_api import Api, ProviderSpec
+
 from llama_stack.core.datatypes import (
     LLAMA_STACK_RUN_CONFIG_VERSION,
     DistributionSpec,
@@ -20,7 +22,6 @@ from llama_stack.core.stack import cast_image_name_to_string, replace_env_vars
 from llama_stack.core.utils.dynamic import instantiate_class_type
 from llama_stack.core.utils.prompt_for_config import prompt_for_config
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import Api, ProviderSpec
 
 logger = get_logger(name=__name__, category="core")
 

src/llama_stack/core/conversations/conversations.py

@@ -8,9 +8,7 @@ import secrets
 import time
 from typing import Any, Literal
 
-from pydantic import BaseModel, TypeAdapter
-
-from llama_stack.apis.conversations.conversations import (
+from llama_stack_api import (
     Conversation,
     ConversationDeletedResource,
     ConversationItem,
@@ -20,6 +18,8 @@ from llama_stack.apis.conversations.conversations import (
     Conversations,
     Metadata,
 )
+from pydantic import BaseModel, TypeAdapter
+
 from llama_stack.core.datatypes import AccessRule, StackRunConfig
 from llama_stack.log import get_logger
 from llama_stack.providers.utils.sqlstore.api import ColumnDefinition, ColumnType

src/llama_stack/core/datatypes.py

@@ -9,22 +9,34 @@ from pathlib import Path
 from typing import Annotated, Any, Literal, Self
 from urllib.parse import urlparse
 
+from llama_stack_api import (
+    Api,
+    Benchmark,
+    BenchmarkInput,
+    Dataset,
+    DatasetInput,
+    DatasetIO,
+    Eval,
+    Inference,
+    Model,
+    ModelInput,
+    ProviderSpec,
+    Resource,
+    Safety,
+    Scoring,
+    ScoringFn,
+    ScoringFnInput,
+    Shield,
+    ShieldInput,
+    ToolGroup,
+    ToolGroupInput,
+    ToolRuntime,
+    VectorIO,
+    VectorStore,
+    VectorStoreInput,
+)
 from pydantic import BaseModel, Field, field_validator, model_validator
 
-from llama_stack.apis.benchmarks import Benchmark, BenchmarkInput
-from llama_stack.apis.datasetio import DatasetIO
-from llama_stack.apis.datasets import Dataset, DatasetInput
-from llama_stack.apis.eval import Eval
-from llama_stack.apis.inference import Inference
-from llama_stack.apis.models import Model, ModelInput
-from llama_stack.apis.resource import Resource
-from llama_stack.apis.safety import Safety
-from llama_stack.apis.scoring import Scoring
-from llama_stack.apis.scoring_functions import ScoringFn, ScoringFnInput
-from llama_stack.apis.shields import Shield, ShieldInput
-from llama_stack.apis.tools import ToolGroup, ToolGroupInput, ToolRuntime
-from llama_stack.apis.vector_io import VectorIO
-from llama_stack.apis.vector_stores import VectorStore, VectorStoreInput
 from llama_stack.core.access_control.datatypes import AccessRule
 from llama_stack.core.storage.datatypes import (
     KVStoreReference,
@@ -32,7 +44,6 @@ from llama_stack.core.storage.datatypes import (
     StorageConfig,
 )
 from llama_stack.log import LoggingConfig
-from llama_stack.providers.datatypes import Api, ProviderSpec
 
 LLAMA_STACK_BUILD_CONFIG_VERSION = 2
 LLAMA_STACK_RUN_CONFIG_VERSION = 2

src/llama_stack/core/distribution.py

@@ -10,17 +10,17 @@ import os
 from typing import Any
 
 import yaml
-from pydantic import BaseModel
-
-from llama_stack.core.datatypes import BuildConfig, DistributionSpec
-from llama_stack.core.external import load_external_apis
-from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import (
+from llama_stack_api import (
     Api,
     InlineProviderSpec,
     ProviderSpec,
     RemoteProviderSpec,
 )
+from pydantic import BaseModel
+
+from llama_stack.core.datatypes import BuildConfig, DistributionSpec
+from llama_stack.core.external import load_external_apis
+from llama_stack.log import get_logger
 
 logger = get_logger(name=__name__, category="core")
 

src/llama_stack/core/external.py

@@ -6,8 +6,8 @@
 
 
 import yaml
+from llama_stack_api import Api, ExternalApiSpec
 
-from llama_stack.apis.datatypes import Api, ExternalApiSpec
 from llama_stack.core.datatypes import BuildConfig, StackRunConfig
 from llama_stack.log import get_logger
 

src/llama_stack/core/inspect.py

@@ -6,19 +6,19 @@
 
 from importlib.metadata import version
 
-from pydantic import BaseModel
-
-from llama_stack.apis.inspect import (
+from llama_stack_api import (
     HealthInfo,
+    HealthStatus,
     Inspect,
     ListRoutesResponse,
     RouteInfo,
     VersionInfo,
 )
+from pydantic import BaseModel
+
 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
-from llama_stack.providers.datatypes import HealthStatus
 
 
 class DistributionInspectConfig(BaseModel):

src/llama_stack/core/library_client.py

@@ -18,6 +18,7 @@ from typing import Any, TypeVar, Union, get_args, get_origin
 import httpx
 import yaml
 from fastapi import Response as FastAPIResponse
+from llama_stack_api import is_unwrapped_body_param
 
 try:
     from llama_stack_client import (
@@ -57,7 +58,6 @@ from llama_stack.core.utils.config import redact_sensitive_fields
 from llama_stack.core.utils.context import preserve_contexts_async_generator
 from llama_stack.core.utils.exec import in_notebook
 from llama_stack.log import get_logger, setup_logging
-from llama_stack.strong_typing.inspection import is_unwrapped_body_param
 
 logger = get_logger(name=__name__, category="core")
 

src/llama_stack/core/prompts/prompts.py

@@ -7,9 +7,9 @@
 import json
 from typing import Any
 
+from llama_stack_api import ListPromptsResponse, Prompt, Prompts
 from pydantic import BaseModel
 
-from llama_stack.apis.prompts import ListPromptsResponse, Prompt, Prompts
 from llama_stack.core.datatypes import StackRunConfig
 from llama_stack.providers.utils.kvstore import KVStore, kvstore_impl
 

src/llama_stack/core/providers.py

@@ -7,11 +7,10 @@
 import asyncio
 from typing import Any
 
+from llama_stack_api import HealthResponse, HealthStatus, ListProvidersResponse, ProviderInfo, Providers
 from pydantic import BaseModel
 
-from llama_stack.apis.providers import ListProvidersResponse, ProviderInfo, Providers
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import HealthResponse, HealthStatus
 
 from .datatypes import StackRunConfig
 from .utils.config import redact_sensitive_fields

src/llama_stack/core/resolver.py

@@ -8,29 +8,46 @@ import importlib.metadata
 import inspect
 from typing import Any
 
-from llama_stack.apis.agents import Agents
-from llama_stack.apis.batches import Batches
-from llama_stack.apis.benchmarks import Benchmarks
-from llama_stack.apis.conversations import Conversations
-from llama_stack.apis.datasetio import DatasetIO
-from llama_stack.apis.datasets import Datasets
-from llama_stack.apis.datatypes import ExternalApiSpec
-from llama_stack.apis.eval import Eval
-from llama_stack.apis.files import Files
-from llama_stack.apis.inference import Inference, InferenceProvider
-from llama_stack.apis.inspect import Inspect
-from llama_stack.apis.models import Models
-from llama_stack.apis.post_training import PostTraining
-from llama_stack.apis.prompts import Prompts
-from llama_stack.apis.providers import Providers as ProvidersAPI
-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 ToolGroups, ToolRuntime
-from llama_stack.apis.vector_io import VectorIO
-from llama_stack.apis.vector_stores import VectorStore
-from llama_stack.apis.version import LLAMA_STACK_API_V1ALPHA
+from llama_stack_api import (
+    LLAMA_STACK_API_V1ALPHA,
+    Agents,
+    Api,
+    Batches,
+    Benchmarks,
+    BenchmarksProtocolPrivate,
+    Conversations,
+    DatasetIO,
+    Datasets,
+    DatasetsProtocolPrivate,
+    Eval,
+    ExternalApiSpec,
+    Files,
+    Inference,
+    InferenceProvider,
+    Inspect,
+    Models,
+    ModelsProtocolPrivate,
+    PostTraining,
+    Prompts,
+    ProviderSpec,
+    RemoteProviderConfig,
+    RemoteProviderSpec,
+    Safety,
+    Scoring,
+    ScoringFunctions,
+    ScoringFunctionsProtocolPrivate,
+    Shields,
+    ShieldsProtocolPrivate,
+    ToolGroups,
+    ToolGroupsProtocolPrivate,
+    ToolRuntime,
+    VectorIO,
+    VectorStore,
+)
+from llama_stack_api import (
+    Providers as ProvidersAPI,
+)
+
 from llama_stack.core.client import get_client_impl
 from llama_stack.core.datatypes import (
     AccessRule,
@@ -44,18 +61,6 @@ from llama_stack.core.external import load_external_apis
 from llama_stack.core.store import DistributionRegistry
 from llama_stack.core.utils.dynamic import instantiate_class_type
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import (
-    Api,
-    BenchmarksProtocolPrivate,
-    DatasetsProtocolPrivate,
-    ModelsProtocolPrivate,
-    ProviderSpec,
-    RemoteProviderConfig,
-    RemoteProviderSpec,
-    ScoringFunctionsProtocolPrivate,
-    ShieldsProtocolPrivate,
-    ToolGroupsProtocolPrivate,
-)
 
 logger = get_logger(name=__name__, category="core")
 

src/llama_stack/core/routers/__init__.py

@@ -6,13 +6,14 @@
 
 from typing import Any
 
+from llama_stack_api import Api, RoutingTable
+
 from llama_stack.core.datatypes import (
     AccessRule,
     RoutedProtocol,
 )
 from llama_stack.core.stack import StackRunConfig
 from llama_stack.core.store import DistributionRegistry
-from llama_stack.providers.datatypes import Api, RoutingTable
 from llama_stack.providers.utils.inference.inference_store import InferenceStore
 
 

src/llama_stack/core/routers/datasets.py

@@ -6,11 +6,9 @@
 
 from typing import Any
 
-from llama_stack.apis.common.responses import PaginatedResponse
-from llama_stack.apis.datasetio import DatasetIO
-from llama_stack.apis.datasets import DatasetPurpose, DataSource
+from llama_stack_api import DatasetIO, DatasetPurpose, DataSource, PaginatedResponse, RoutingTable
+
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import RoutingTable
 
 logger = get_logger(name=__name__, category="core::routers")
 

src/llama_stack/core/routers/eval_scoring.py

@@ -6,15 +6,19 @@
 
 from typing import Any
 
-from llama_stack.apis.eval import BenchmarkConfig, Eval, EvaluateResponse, Job
-from llama_stack.apis.scoring import (
+from llama_stack_api import (
+    BenchmarkConfig,
+    Eval,
+    EvaluateResponse,
+    Job,
+    RoutingTable,
     ScoreBatchResponse,
     ScoreResponse,
     Scoring,
     ScoringFnParams,
 )
+
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import RoutingTable
 
 logger = get_logger(name=__name__, category="core::routers")
 

src/llama_stack/core/routers/inference.py

@@ -11,17 +11,19 @@ from datetime import UTC, datetime
 from typing import Annotated, Any
 
 from fastapi import Body
-from openai.types.chat import ChatCompletionToolChoiceOptionParam as OpenAIChatCompletionToolChoiceOptionParam
-from openai.types.chat import ChatCompletionToolParam as OpenAIChatCompletionToolParam
-from pydantic import TypeAdapter
-
-from llama_stack.apis.common.errors import ModelNotFoundError, ModelTypeError
-from llama_stack.apis.inference import (
+from llama_stack_api import (
+    HealthResponse,
+    HealthStatus,
     Inference,
     ListOpenAIChatCompletionResponse,
+    ModelNotFoundError,
+    ModelType,
+    ModelTypeError,
     OpenAIAssistantMessageParam,
     OpenAIChatCompletion,
     OpenAIChatCompletionChunk,
+    OpenAIChatCompletionContentPartImageParam,
+    OpenAIChatCompletionContentPartTextParam,
     OpenAIChatCompletionRequestWithExtraBody,
     OpenAIChatCompletionToolCall,
     OpenAIChatCompletionToolCallFunction,
@@ -35,18 +37,17 @@ from llama_stack.apis.inference import (
     OpenAIMessageParam,
     Order,
     RerankResponse,
+    RoutingTable,
 )
-from llama_stack.apis.inference.inference import (
-    OpenAIChatCompletionContentPartImageParam,
-    OpenAIChatCompletionContentPartTextParam,
-)
-from llama_stack.apis.models import ModelType
+from openai.types.chat import ChatCompletionToolChoiceOptionParam as OpenAIChatCompletionToolChoiceOptionParam
+from openai.types.chat import ChatCompletionToolParam as OpenAIChatCompletionToolParam
+from pydantic import TypeAdapter
+
 from llama_stack.core.telemetry.telemetry import MetricEvent
 from llama_stack.core.telemetry.tracing import enqueue_event, get_current_span
 from llama_stack.log import get_logger
 from llama_stack.models.llama.llama3.chat_format import ChatFormat
 from llama_stack.models.llama.llama3.tokenizer import Tokenizer
-from llama_stack.providers.datatypes import HealthResponse, HealthStatus, RoutingTable
 from llama_stack.providers.utils.inference.inference_store import InferenceStore
 
 logger = get_logger(name=__name__, category="core::routers")

src/llama_stack/core/routers/safety.py

@@ -6,13 +6,10 @@
 
 from typing import Any
 
-from llama_stack.apis.inference import OpenAIMessageParam
-from llama_stack.apis.safety import RunShieldResponse, Safety
-from llama_stack.apis.safety.safety import ModerationObject
-from llama_stack.apis.shields import Shield
+from llama_stack_api import ModerationObject, OpenAIMessageParam, RoutingTable, RunShieldResponse, Safety, Shield
+
 from llama_stack.core.datatypes import SafetyConfig
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import RoutingTable
 
 logger = get_logger(name=__name__, category="core::routers")
 

src/llama_stack/core/routers/tool_runtime.py

@@ -6,13 +6,12 @@
 
 from typing import Any
 
-from llama_stack.apis.common.content_types import (
+from llama_stack_api import (
     URL,
-)
-from llama_stack.apis.tools import (
     ListToolDefsResponse,
     ToolRuntime,
 )
+
 from llama_stack.log import get_logger
 
 from ..routing_tables.toolgroups import ToolGroupsRoutingTable

src/llama_stack/core/routers/vector_io.py

@@ -9,14 +9,16 @@ import uuid
 from typing import Annotated, Any
 
 from fastapi import Body
-
-from llama_stack.apis.common.content_types import InterleavedContent
-from llama_stack.apis.models import ModelType
-from llama_stack.apis.vector_io import (
+from llama_stack_api import (
     Chunk,
+    HealthResponse,
+    HealthStatus,
+    InterleavedContent,
+    ModelType,
     OpenAICreateVectorStoreFileBatchRequestWithExtraBody,
     OpenAICreateVectorStoreRequestWithExtraBody,
     QueryChunksResponse,
+    RoutingTable,
     SearchRankingOptions,
     VectorIO,
     VectorStoreChunkingStrategy,
@@ -33,9 +35,9 @@ from llama_stack.apis.vector_io import (
     VectorStoreObject,
     VectorStoreSearchResponsePage,
 )
+
 from llama_stack.core.datatypes import VectorStoresConfig
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import HealthResponse, HealthStatus, RoutingTable
 
 logger = get_logger(name=__name__, category="core::routers")
 

src/llama_stack/core/routing_tables/benchmarks.py

@@ -6,7 +6,8 @@
 
 from typing import Any
 
-from llama_stack.apis.benchmarks import Benchmark, Benchmarks, ListBenchmarksResponse
+from llama_stack_api import Benchmark, Benchmarks, ListBenchmarksResponse
+
 from llama_stack.core.datatypes import (
     BenchmarkWithOwner,
 )

src/llama_stack/core/routing_tables/common.py

@@ -6,9 +6,8 @@
 
 from typing import Any
 
-from llama_stack.apis.common.errors import ModelNotFoundError
-from llama_stack.apis.models import Model
-from llama_stack.apis.resource import ResourceType
+from llama_stack_api import Api, Model, ModelNotFoundError, ResourceType, RoutingTable
+
 from llama_stack.core.access_control.access_control import AccessDeniedError, is_action_allowed
 from llama_stack.core.access_control.datatypes import Action
 from llama_stack.core.datatypes import (
@@ -21,7 +20,6 @@ from llama_stack.core.datatypes import (
 from llama_stack.core.request_headers import get_authenticated_user
 from llama_stack.core.store import DistributionRegistry
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import Api, RoutingTable
 
 logger = get_logger(name=__name__, category="core::routing_tables")
 

src/llama_stack/core/routing_tables/datasets.py

@@ -7,18 +7,19 @@
 import uuid
 from typing import Any
 
-from llama_stack.apis.common.errors import DatasetNotFoundError
-from llama_stack.apis.datasets import (
+from llama_stack_api import (
     Dataset,
+    DatasetNotFoundError,
     DatasetPurpose,
     Datasets,
     DatasetType,
     DataSource,
     ListDatasetsResponse,
+    ResourceType,
     RowsDataSource,
     URIDataSource,
 )
-from llama_stack.apis.resource import ResourceType
+
 from llama_stack.core.datatypes import (
     DatasetWithOwner,
 )

src/llama_stack/core/routing_tables/models.py

@@ -7,8 +7,16 @@
 import time
 from typing import Any
 
-from llama_stack.apis.common.errors import ModelNotFoundError
-from llama_stack.apis.models import ListModelsResponse, Model, Models, ModelType, OpenAIListModelsResponse, OpenAIModel
+from llama_stack_api import (
+    ListModelsResponse,
+    Model,
+    ModelNotFoundError,
+    Models,
+    ModelType,
+    OpenAIListModelsResponse,
+    OpenAIModel,
+)
+
 from llama_stack.core.datatypes import (
     ModelWithOwner,
     RegistryEntrySource,

src/llama_stack/core/routing_tables/scoring_functions.py

@@ -4,14 +4,15 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
-from llama_stack.apis.common.type_system import ParamType
-from llama_stack.apis.resource import ResourceType
-from llama_stack.apis.scoring_functions import (
+from llama_stack_api import (
     ListScoringFunctionsResponse,
+    ParamType,
+    ResourceType,
     ScoringFn,
     ScoringFnParams,
     ScoringFunctions,
 )
+
 from llama_stack.core.datatypes import (
     ScoringFnWithOwner,
 )

src/llama_stack/core/routing_tables/shields.py

@@ -6,8 +6,8 @@
 
 from typing import Any
 
-from llama_stack.apis.resource import ResourceType
-from llama_stack.apis.shields import ListShieldsResponse, Shield, Shields
+from llama_stack_api import ListShieldsResponse, ResourceType, Shield, Shields
+
 from llama_stack.core.datatypes import (
     ShieldWithOwner,
 )

src/llama_stack/core/routing_tables/toolgroups.py

@@ -6,9 +6,16 @@
 
 from typing import Any
 
-from llama_stack.apis.common.content_types import URL
-from llama_stack.apis.common.errors import ToolGroupNotFoundError
-from llama_stack.apis.tools import ListToolDefsResponse, ListToolGroupsResponse, ToolDef, ToolGroup, ToolGroups
+from llama_stack_api import (
+    URL,
+    ListToolDefsResponse,
+    ListToolGroupsResponse,
+    ToolDef,
+    ToolGroup,
+    ToolGroupNotFoundError,
+    ToolGroups,
+)
+
 from llama_stack.core.datatypes import AuthenticationRequiredError, ToolGroupWithOwner
 from llama_stack.log import get_logger
 

src/llama_stack/core/routing_tables/vector_stores.py

@@ -6,12 +6,12 @@
 
 from typing import Any
 
-from llama_stack.apis.common.errors import ModelNotFoundError, ModelTypeError
-from llama_stack.apis.models import ModelType
-from llama_stack.apis.resource import ResourceType
-
 # Removed VectorStores import to avoid exposing public API
-from llama_stack.apis.vector_io.vector_io import (
+from llama_stack_api import (
+    ModelNotFoundError,
+    ModelType,
+    ModelTypeError,
+    ResourceType,
     SearchRankingOptions,
     VectorStoreChunkingStrategy,
     VectorStoreDeleteResponse,
@@ -22,6 +22,7 @@ from llama_stack.apis.vector_io.vector_io import (
     VectorStoreObject,
     VectorStoreSearchResponsePage,
 )
+
 from llama_stack.core.datatypes import (
     VectorStoreWithOwner,
 )

src/llama_stack/core/server/auth_providers.py

@@ -11,9 +11,9 @@ from urllib.parse import parse_qs, urljoin, urlparse
 
 import httpx
 import jwt
+from llama_stack_api import TokenValidationError
 from pydantic import BaseModel, Field
 
-from llama_stack.apis.common.errors import TokenValidationError
 from llama_stack.core.datatypes import (
     AuthenticationConfig,
     CustomAuthConfig,

src/llama_stack/core/server/routes.py

@@ -10,11 +10,10 @@ from collections.abc import Callable
 from typing import Any
 
 from aiohttp import hdrs
+from llama_stack_api import Api, ExternalApiSpec, WebMethod
 from starlette.routing import Route
 
-from llama_stack.apis.datatypes import Api, ExternalApiSpec
 from llama_stack.core.resolver import api_protocol_map
-from llama_stack.schema_utils import WebMethod
 
 EndpointFunc = Callable[..., Any]
 PathParams = dict[str, str]

src/llama_stack/core/server/server.py

@@ -28,11 +28,10 @@ from fastapi import Path as FastapiPath
 from fastapi.exceptions import RequestValidationError
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import JSONResponse, StreamingResponse
+from llama_stack_api import Api, ConflictError, PaginatedResponse, ResourceNotFoundError
 from openai import BadRequestError
 from pydantic import BaseModel, ValidationError
 
-from llama_stack.apis.common.errors import ConflictError, ResourceNotFoundError
-from llama_stack.apis.common.responses import PaginatedResponse
 from llama_stack.core.access_control.access_control import AccessDeniedError
 from llama_stack.core.datatypes import (
     AuthenticationRequiredError,
@@ -58,7 +57,6 @@ from llama_stack.core.utils.config import redact_sensitive_fields
 from llama_stack.core.utils.config_resolution import Mode, resolve_config_or_distro
 from llama_stack.core.utils.context import preserve_contexts_async_generator
 from llama_stack.log import LoggingConfig, get_logger, setup_logging
-from llama_stack.providers.datatypes import Api
 
 from .auth import AuthenticationMiddleware
 from .quota import QuotaMiddleware

src/llama_stack/core/stack.py

@@ -12,27 +12,31 @@ import tempfile
 from typing import Any
 
 import yaml
+from llama_stack_api import (
+    Agents,
+    Api,
+    Batches,
+    Benchmarks,
+    Conversations,
+    DatasetIO,
+    Datasets,
+    Eval,
+    Files,
+    Inference,
+    Inspect,
+    Models,
+    PostTraining,
+    Prompts,
+    Providers,
+    Safety,
+    Scoring,
+    ScoringFunctions,
+    Shields,
+    ToolGroups,
+    ToolRuntime,
+    VectorIO,
+)
 
-from llama_stack.apis.agents import Agents
-from llama_stack.apis.batches import Batches
-from llama_stack.apis.benchmarks import Benchmarks
-from llama_stack.apis.conversations import Conversations
-from llama_stack.apis.datasetio import DatasetIO
-from llama_stack.apis.datasets import Datasets
-from llama_stack.apis.eval import Eval
-from llama_stack.apis.files import Files
-from llama_stack.apis.inference import Inference
-from llama_stack.apis.inspect import Inspect
-from llama_stack.apis.models import Models
-from llama_stack.apis.post_training import PostTraining
-from llama_stack.apis.prompts import Prompts
-from llama_stack.apis.providers import Providers
-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 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
 from llama_stack.core.distribution import get_provider_registry
@@ -54,7 +58,6 @@ from llama_stack.core.storage.datatypes import (
 from llama_stack.core.store.registry import create_dist_registry
 from llama_stack.core.utils.dynamic import instantiate_class_type
 from llama_stack.log import get_logger
-from llama_stack.providers.datatypes import Api
 
 logger = get_logger(name=__name__, category="core")
 

src/llama_stack/core/telemetry/telemetry.py

@@ -16,6 +16,7 @@ from typing import (
     cast,
 )
 
+from llama_stack_api import json_schema_type, register_schema
 from opentelemetry import metrics, trace
 from opentelemetry.exporter.otlp.proto.http.metric_exporter import OTLPMetricExporter
 from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
@@ -28,7 +29,6 @@ from pydantic import BaseModel, Field
 
 from llama_stack.log import get_logger
 from llama_stack.models.llama.datatypes import Primitive
-from llama_stack.schema_utils import json_schema_type, register_schema
 
 ROOT_SPAN_MARKERS = ["__root__", "__root_span__"]