Note

Detect and handle null/empty LLM responses as context-window overflows, triggering summarization; update typings to BaseAgent and add focused tests.

• Agents/Execution:
  • Treat null/empty LLM responses as potential context-window overflows via new is_null_response_because_context_length_exceeded and route to handle_context_length in CrewAgentExecutor and LiteAgent.
  • Add LLMMessage typing where needed.
• Utilities:
  • Add is_null_response_because_context_length_exceeded(exception, messages, llm) in utilities/agent_utils.py.
  • Broaden is_context_length_exceeded docstring/signature; no behavioral regression.
• Types/API:
  • Use BaseAgent in TaskEvaluator and update related imports/usages (e.g., tool utils typing).
• Tests:
  • Add unit tests for null/empty response handling and new utility, plus executor integration test ensuring summarization path and final answer.

^{Written by Cursor Bugbot for commit da3f195. This will update automatically on new commits. Configure here.}