refactor: Add ProviderContext for a flexible storage directory

- Introduce ProviderContext class to decouple provider storage paths from absolute paths
- Add storage_dir attribute to StackRunConfig to accept CLI options
- Implement storage directory resolution with prioritized fallbacks:
  1. CLI option (--state-directory)
  2. Environment variable (LLAMA_STACK_STATE_DIR)
  3. Default distribution directory
- Standardize provider signatures to follow context, config, deps pattern
- Update provider implementations to use the new context-based approach
- Add comprehensive tests to verify state directory resolution

llama_stack/providers/inline/scoring/basic/__init__.py

@@ -6,17 +6,20 @@
 from typing import Any
 
 from llama_stack.distribution.datatypes import Api
+from llama_stack.providers.datatypes import ProviderContext
 
 from .config import BasicScoringConfig
 
 
 async def get_provider_impl(
+    context: ProviderContext,
     config: BasicScoringConfig,
     deps: dict[Api, Any],
 ):
     from .scoring import BasicScoringImpl
 
     impl = BasicScoringImpl(
+        context,
         config,
         deps[Api.datasetio],
         deps[Api.datasets],

llama_stack/providers/inline/scoring/basic/scoring.py

@@ -15,7 +15,7 @@ from llama_stack.apis.scoring import (
 )
 from llama_stack.apis.scoring_functions import ScoringFn, ScoringFnParams
 from llama_stack.distribution.datatypes import Api
-from llama_stack.providers.datatypes import ScoringFunctionsProtocolPrivate
+from llama_stack.providers.datatypes import ProviderContext, ScoringFunctionsProtocolPrivate
 from llama_stack.providers.utils.common.data_schema_validator import (
     get_valid_schemas,
     validate_dataset_schema,
@@ -49,10 +49,12 @@ class BasicScoringImpl(
 ):
     def __init__(
         self,
+        context: ProviderContext,
         config: BasicScoringConfig,
         datasetio_api: DatasetIO,
         datasets_api: Datasets,
     ) -> None:
+        self.context = context
         self.config = config
         self.datasetio_api = datasetio_api
         self.datasets_api = datasets_api

llama_stack/providers/inline/scoring/braintrust/__init__.py

@@ -8,6 +8,7 @@ from typing import Any
 from pydantic import BaseModel
 
 from llama_stack.distribution.datatypes import Api
+from llama_stack.providers.datatypes import ProviderContext
 
 from .config import BraintrustScoringConfig
 
@@ -17,6 +18,7 @@ class BraintrustProviderDataValidator(BaseModel):
 
 
 async def get_provider_impl(
+    context: ProviderContext,
     config: BraintrustScoringConfig,
     deps: dict[Api, Any],
 ):

llama_stack/providers/inline/scoring/llm_as_judge/__init__.py

@@ -6,11 +6,13 @@
 from typing import Any
 
 from llama_stack.distribution.datatypes import Api
+from llama_stack.providers.datatypes import ProviderContext
 
 from .config import LlmAsJudgeScoringConfig
 
 
 async def get_provider_impl(
+    context: ProviderContext,
     config: LlmAsJudgeScoringConfig,
     deps: dict[Api, Any],
 ):