- Create LLMQuotaLimitExceededException following CrewAI's existing pattern
- Add quota limit error handling in both streaming and non-streaming LLM calls
- Update error handling in agent execution and crew agent executor
- Add comprehensive tests for quota limit scenarios
- Fixes issue #3434: Handle RateLimitError gracefully instead of crashing
The implementation catches litellm.exceptions.RateLimitError and converts it
to a CrewAI-specific exception, allowing tasks to detect quota limits and
shut down gracefully instead of crashing with unhandled exceptions.
Co-Authored-By: João <joao@crewai.com>
c763457e8d