Fix additional linter errors

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

Author: jgarciao

tests/rag/test_rag.py

@@ -57,7 +57,8 @@ def test_rag_basic_inference(self, lls_client: LlamaStackClient) -> None:
         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()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
@@ -84,7 +85,8 @@ def test_rag_simple_agent(self, lls_client: LlamaStackClient) -> None:
         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()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
@@ -120,7 +122,8 @@ def test_rag_build_rag_agent(self, lls_client: LlamaStackClient) -> None:
         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()
         model_id = next(m for m in models if m.api_model_type == "llm").identifier
@@ -164,7 +167,7 @@ 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={},
                 )

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})