feat!: providers use unified 'persistence' field

BREAKING CHANGE: Provider config field names changed for semantic clarity

- Rename kvstore → persistence for all providers
- Simple providers: flat persistence with backend reference
- Complex providers (agents): nested persistence.agent_state + persistence.responses
- Files provider: metadata_store → persistence
- Provider configs now clearly express 'how do I persist?' not 'what type of store?'

Example:
  # Before
  config:
    kvstore:
      backend: kvstore
      namespace: faiss

  # After
  config:
    persistence:
      backend: kvstore
      namespace: faiss

  # Agents (nested)
  config:
    persistence:
      agent_state:
        backend: kvstore
        namespace: agents
      responses:
        backend: sqlstore
        namespace: responses

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

@@ -77,8 +77,8 @@ class MetaReferenceAgentsImpl(Agents):
         self.policy = policy
 
     async def initialize(self) -> None:
-        self.persistence_store = await kvstore_impl(self.config.persistence_store)
-        self.responses_store = ResponsesStore(self.config.responses_store, self.policy)
+        self.persistence_store = await kvstore_impl(self.config.persistence.agent_state)
+        self.responses_store = ResponsesStore(self.config.persistence.responses, self.policy)
         await self.responses_store.initialize()
         self.openai_responses_impl = OpenAIResponsesImpl(
             inference_api=self.inference_api,