Fix additional linter errors

Signed-off-by: Jorge Garcia Oncins <jgarciao@redhat.com>

Author: jgarciao

tests/rag/conftest.py

@@ -127,7 +127,7 @@ def wait_for_llama_stack_ready(client: LlamaStackClient) -> bool:
 
 
 @pytest.fixture(scope="class")
-def lls_client(
+def rag_lls_client(
     admin_client: DynamicClient,
     rag_test_namespace: Namespace | Project,
     llama_stack_distribution_deployment: Deployment,
@@ -138,7 +138,7 @@ def lls_client(
 
     Args:
         admin_client (DynamicClient): Kubernetes dynamic client for cluster operations
-        rag_test_namespace (Namespace): Namespace or project containing RAG test resources
+        rag_test_namespace (Namespace | Project): Namespace or project containing RAG test resources
         llama_stack_distribution_deployment (Deployment): LlamaStack distribution deployment resource
 
     Yields:
@@ -156,7 +156,7 @@ def lls_client(
                 base_url="http://localhost:8321",
                 timeout=120.0,
             )
-            wait_for_llama_stack_ready(client)
+            wait_for_llama_stack_ready(client=client)
             yield client
     except Exception as e:
         LOGGER.error(f"Failed to set up port forwarding: {e}")

tests/rag/test_rag.py

@@ -21,7 +21,7 @@ class TestRag:
 
     @pytest.mark.smoke
     def test_llama_stack_server(
-        self, llama_stack_distribution_deployment: Deployment, lls_client: LlamaStackClient
+        self, llama_stack_distribution_deployment: Deployment, rag_lls_client: LlamaStackClient
     ) -> None:
         """
         Test LlamaStack Server deployment and verify required models are available.
@@ -33,7 +33,7 @@ def test_llama_stack_server(
         """
         llama_stack_distribution_deployment.wait_for_replicas()
 
-        models = lls_client.models.list()
+        models = rag_lls_client.models.list()
         assert models is not None, "No models returned from LlamaStackClient"
 
         llm_model = next((m for m in models if m.api_model_type == "llm"), None)
@@ -50,19 +50,20 @@ def test_llama_stack_server(
         assert embedding_dimension is not None, "No embedding_dimension set in embedding model"
 
     @pytest.mark.smoke
-    def test_rag_basic_inference(self, lls_client: LlamaStackClient) -> None:
+    def test_rag_basic_inference(self, rag_lls_client: LlamaStackClient) -> None:
         """
         Test basic chat completion inference through LlamaStack client.
 
         Validates that the server can perform text generation using the chat completions API
         and provides factually correct responses.
 
-        Based on the example available at https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
+        Based on the example available at
+        https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
         """
-        models = lls_client.models.list()
+        models = rag_lls_client.models.list()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
 
-        response = lls_client.chat.completions.create(
+        response = rag_lls_client.chat.completions.create(
             model=model_id,
             messages=[
                 {"role": "system", "content": "You are a helpful assistant."},
@@ -77,18 +78,19 @@ def test_rag_basic_inference(self, lls_client: LlamaStackClient) -> None:
         assert "Paris" in content, "The LLM didn't provide the expected answer to the prompt"
 
     @pytest.mark.smoke
-    def test_rag_simple_agent(self, lls_client: LlamaStackClient) -> None:
+    def test_rag_simple_agent(self, rag_lls_client: LlamaStackClient) -> None:
         """
         Test basic agent creation and conversation capabilities.
 
         Validates agent creation, session management, and turn-based interactions
         with both identity and capability questions.
 
-        Based on the example available at https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
+        Based on the example available at
+        https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
         """
-        models = lls_client.models.list()
+        models = rag_lls_client.models.list()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
-        agent = Agent(lls_client, model=model_id, instructions="You are a helpful assistant.")
+        agent = Agent(client=rag_lls_client, model=model_id, instructions="You are a helpful assistant.")
         s_id = agent.create_session(session_name=f"s{uuid.uuid4().hex}")
 
         # Test identity question
@@ -112,17 +114,18 @@ def test_rag_simple_agent(self, lls_client: LlamaStackClient) -> None:
         assert "answers" in content, "The LLM didn't provide the expected answer to the prompt"
 
     @pytest.mark.smoke
-    def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
+    def test_rag_build_rag_agent(self, rag_lls_client: LlamaStackClient) -> None:
         """
         Test full RAG pipeline with vector database integration and knowledge retrieval.
 
         Creates a RAG agent with PyTorch torchtune documentation, tests knowledge queries
         about fine-tuning techniques (LoRA, QAT, memory optimizations), and validates
         that responses contain expected technical keywords.
 
-        Based on the example available at https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
+        Based on the example available at
+        https://llama-stack.readthedocs.io/en/latest/getting_started/detailed_tutorial.html#step-4-run-the-demos
         """
-        models = lls_client.models.list()
+        models = rag_lls_client.models.list()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
         embedding_model = next(m for m in models if m.api_model_type == "embedding")
 
@@ -131,7 +134,7 @@ def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
         # Create a vector database instance
         vector_db_id = f"v{uuid.uuid4().hex}"
 
-        lls_client.vector_dbs.register(
+        rag_lls_client.vector_dbs.register(
             vector_db_id=vector_db_id,
             embedding_model=embedding_model.identifier,
             embedding_dimension=embedding_dimension,
@@ -141,7 +144,7 @@ def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
         try:
             # Create the RAG agent connected to the vector database
             rag_agent = Agent(
-                lls_client,
+                client=rag_lls_client,
                 model=model_id,
                 instructions="You are a helpful assistant. Use the RAG tool to answer questions as needed.",
                 tools=[
@@ -164,14 +167,14 @@ def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
             documents = [
                 RAGDocument(
                     document_id=f"num-{i}",
-                    content=f"https://raw.githubusercontent.com/pytorch/torchtune/refs/tags/v0.6.1/docs/source/tutorials/{url}",
+                    content=f"https://raw.githubusercontent.com/pytorch/torchtune/refs/tags/v0.6.1/docs/source/tutorials/{url}", # noqa
                     mime_type="text/plain",
                     metadata={},
                 )
                 for i, url in enumerate(urls)
             ]
 
-            lls_client.tool_runtime.rag_tool.insert(
+            rag_lls_client.tool_runtime.rag_tool.insert(
                 documents=documents,
                 vector_db_id=vector_db_id,
                 chunk_size_in_tokens=512,
@@ -243,6 +246,6 @@ def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
         finally:
             # Cleanup: unregister the vector database to prevent resource leaks
             try:
-                lls_client.vector_dbs.unregister(vector_db_id)
+                rag_lls_client.vector_dbs.unregister(vector_db_id)
             except Exception as e:
                 LOGGER.warning(f"Failed to unregister vector database {vector_db_id}: {e}")

utilities/rag_utils.py

@@ -1,7 +1,7 @@
 from contextlib import contextmanager
 from ocp_resources.resource import NamespacedResource
 from kubernetes.dynamic import DynamicClient
-from typing import Any, Dict, Generator, List, TypedDict
+from typing import Any, Dict, Generator, List, TypedDict, cast
 from llama_stack_client import Agent, AgentEventLogger
 from simple_logger.logger import get_logger
 
@@ -220,8 +220,8 @@ def validate_rag_agent_responses(
         "successful_turns": successful_turns,
         "failed_turns": total_turns - successful_turns,
         "success_rate": successful_turns / total_turns if total_turns > 0 else 0,
-        "total_events": sum(result["event_count"] for result in all_results),
-        "total_response_length": sum(result["response_length"] for result in all_results),
+        "total_events": sum(cast(TurnResult, result)["event_count"] for result in all_results),
+        "total_response_length": sum(cast(TurnResult, result)["response_length"] for result in all_results),
     }
 
     overall_success = successful_turns == total_turns
@@ -235,4 +235,4 @@ def validate_rag_agent_responses(
         LOGGER.info(f"Success rate: {summary['success_rate']:.1%}")
         LOGGER.info(f"Overall result: {'✓ PASSED' if overall_success else '✗ FAILED'}")
 
-    return {"success": overall_success, "results": all_results, "summary": summary}
+    return cast(ValidationResult, {"success": overall_success, "results": all_results, "summary": summary})