Address review feedback: Improve UTC compatibility implementation

- Enhance datetime_compat.py documentation
- Add edge case tests for UTC timezone handling
- Update mock patching in tests
- Fix utcnow usage in sqlite.py

Co-Authored-By: Joe Moura <joao@crewai.com>

src/crewai/agent.py

@@ -19,25 +19,17 @@ from crewai.tools.agent_tools.agent_tools import AgentTools
 from crewai.utilities import Converter, Prompts
 from crewai.utilities.constants import TRAINED_AGENTS_DATA_FILE, TRAINING_DATA_FILE
 from crewai.utilities.converter import generate_model_description
+from crewai.utilities.events.agent_events import (
+    AgentExecutionCompletedEvent,
+    AgentExecutionErrorEvent,
+    AgentExecutionStartedEvent,
+)
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
 from crewai.utilities.llm_utils import create_llm
 from crewai.utilities.token_counter_callback import TokenCalcHandler
 from crewai.utilities.training_handler import CrewTrainingHandler
 
-agentops = None
 
-try:
-    import agentops  # type: ignore # Name "agentops" is already defined
-    from agentops import track_agent  # type: ignore
-except ImportError:
-
-    def track_agent():
-        def noop(f):
-            return f
-
-        return noop
-
-
-@track_agent()
 class Agent(BaseAgent):
     """Represents an agent in a system.
 
@@ -240,6 +232,15 @@ class Agent(BaseAgent):
             task_prompt = self._use_trained_data(task_prompt=task_prompt)
 
         try:
+            crewai_event_bus.emit(
+                self,
+                event=AgentExecutionStartedEvent(
+                    agent=self,
+                    tools=self.tools,
+                    task_prompt=task_prompt,
+                    task=task,
+                ),
+            )
             result = self.agent_executor.invoke(
                 {
                     "input": task_prompt,
@@ -251,9 +252,25 @@ class Agent(BaseAgent):
         except Exception as e:
             if e.__class__.__module__.startswith("litellm"):
                 # Do not retry on litellm errors
+                crewai_event_bus.emit(
+                    self,
+                    event=AgentExecutionErrorEvent(
+                        agent=self,
+                        task=task,
+                        error=str(e),
+                    ),
+                )
                 raise e
             self._times_executed += 1
             if self._times_executed > self.max_retry_limit:
+                crewai_event_bus.emit(
+                    self,
+                    event=AgentExecutionErrorEvent(
+                        agent=self,
+                        task=task,
+                        error=str(e),
+                    ),
+                )
                 raise e
             result = self.execute_task(task, context, tools)
 
@@ -266,7 +283,10 @@ class Agent(BaseAgent):
         for tool_result in self.tools_results:  # type: ignore # Item "None" of "list[Any] | None" has no attribute "__iter__" (not iterable)
             if tool_result.get("result_as_answer", False):
                 result = tool_result["result"]
-
+        crewai_event_bus.emit(
+            self,
+            event=AgentExecutionCompletedEvent(agent=self, task=task, output=result),
+        )
         return result
 
     def create_agent_executor(

src/crewai/agents/agent_builder/base_agent.py

@@ -20,8 +20,7 @@ from crewai.agents.cache.cache_handler import CacheHandler
 from crewai.agents.tools_handler import ToolsHandler
 from crewai.knowledge.knowledge import Knowledge
 from crewai.knowledge.source.base_knowledge_source import BaseKnowledgeSource
-from crewai.tools import BaseTool
-from crewai.tools.base_tool import Tool
+from crewai.tools.base_tool import BaseTool, Tool
 from crewai.utilities import I18N, Logger, RPMController
 from crewai.utilities.config import process_config
 from crewai.utilities.converter import Converter
@@ -112,7 +111,7 @@ class BaseAgent(ABC, BaseModel):
         default=False,
         description="Enable agent to delegate and ask questions among each other.",
     )
-    tools: Optional[List[Any]] = Field(
+    tools: Optional[List[BaseTool]] = Field(
         default_factory=list, description="Tools at agents' disposal"
     )
     max_iter: int = Field(

src/crewai/agents/crew_agent_executor.py

@@ -18,6 +18,12 @@ from crewai.tools.base_tool import BaseTool
 from crewai.tools.tool_usage import ToolUsage, ToolUsageErrorException
 from crewai.utilities import I18N, Printer
 from crewai.utilities.constants import MAX_LLM_RETRY, TRAINING_DATA_FILE
+from crewai.utilities.events import (
+    ToolUsageErrorEvent,
+    ToolUsageStartedEvent,
+    crewai_event_bus,
+)
+from crewai.utilities.events.tool_usage_events import ToolUsageStartedEvent
 from crewai.utilities.exceptions.context_window_exceeding_exception import (
     LLMContextLengthExceededException,
 )
@@ -107,11 +113,11 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
             )
             raise
         except Exception as e:
+            self._handle_unknown_error(e)
             if e.__class__.__module__.startswith("litellm"):
                 # Do not retry on litellm errors
                 raise e
             else:
-                self._handle_unknown_error(e)
                 raise e
 
         if self.ask_for_human_input:
@@ -349,40 +355,68 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
                 )
 
     def _execute_tool_and_check_finality(self, agent_action: AgentAction) -> ToolResult:
-        tool_usage = ToolUsage(
-            tools_handler=self.tools_handler,
-            tools=self.tools,
-            original_tools=self.original_tools,
-            tools_description=self.tools_description,
-            tools_names=self.tools_names,
-            function_calling_llm=self.function_calling_llm,
-            task=self.task,  # type: ignore[arg-type]
-            agent=self.agent,
-            action=agent_action,
-        )
-        tool_calling = tool_usage.parse_tool_calling(agent_action.text)
-
-        if isinstance(tool_calling, ToolUsageErrorException):
-            tool_result = tool_calling.message
-            return ToolResult(result=tool_result, result_as_answer=False)
-        else:
-            if tool_calling.tool_name.casefold().strip() in [
-                name.casefold().strip() for name in self.tool_name_to_tool_map
-            ] or tool_calling.tool_name.casefold().replace("_", " ") in [
-                name.casefold().strip() for name in self.tool_name_to_tool_map
-            ]:
-                tool_result = tool_usage.use(tool_calling, agent_action.text)
-                tool = self.tool_name_to_tool_map.get(tool_calling.tool_name)
-                if tool:
-                    return ToolResult(
-                        result=tool_result, result_as_answer=tool.result_as_answer
-                    )
-            else:
-                tool_result = self._i18n.errors("wrong_tool_name").format(
-                    tool=tool_calling.tool_name,
-                    tools=", ".join([tool.name.casefold() for tool in self.tools]),
+        try:
+            if self.agent:
+                crewai_event_bus.emit(
+                    self,
+                    event=ToolUsageStartedEvent(
+                        agent_key=self.agent.key,
+                        agent_role=self.agent.role,
+                        tool_name=agent_action.tool,
+                        tool_args=agent_action.tool_input,
+                        tool_class=agent_action.tool,
+                    ),
                 )
-        return ToolResult(result=tool_result, result_as_answer=False)
+            tool_usage = ToolUsage(
+                tools_handler=self.tools_handler,
+                tools=self.tools,
+                original_tools=self.original_tools,
+                tools_description=self.tools_description,
+                tools_names=self.tools_names,
+                function_calling_llm=self.function_calling_llm,
+                task=self.task,  # type: ignore[arg-type]
+                agent=self.agent,
+                action=agent_action,
+            )
+            tool_calling = tool_usage.parse_tool_calling(agent_action.text)
+
+            if isinstance(tool_calling, ToolUsageErrorException):
+                tool_result = tool_calling.message
+                return ToolResult(result=tool_result, result_as_answer=False)
+            else:
+                if tool_calling.tool_name.casefold().strip() in [
+                    name.casefold().strip() for name in self.tool_name_to_tool_map
+                ] or tool_calling.tool_name.casefold().replace("_", " ") in [
+                    name.casefold().strip() for name in self.tool_name_to_tool_map
+                ]:
+                    tool_result = tool_usage.use(tool_calling, agent_action.text)
+                    tool = self.tool_name_to_tool_map.get(tool_calling.tool_name)
+                    if tool:
+                        return ToolResult(
+                            result=tool_result, result_as_answer=tool.result_as_answer
+                        )
+                else:
+                    tool_result = self._i18n.errors("wrong_tool_name").format(
+                        tool=tool_calling.tool_name,
+                        tools=", ".join([tool.name.casefold() for tool in self.tools]),
+                    )
+                return ToolResult(result=tool_result, result_as_answer=False)
+
+        except Exception as e:
+            # TODO: drop
+            if self.agent:
+                crewai_event_bus.emit(
+                    self,
+                    event=ToolUsageErrorEvent(  # validation error
+                        agent_key=self.agent.key,
+                        agent_role=self.agent.role,
+                        tool_name=agent_action.tool,
+                        tool_args=agent_action.tool_input,
+                        tool_class=agent_action.tool,
+                        error=str(e),
+                    ),
+                )
+            raise e
 
     def _summarize_messages(self) -> None:
         messages_groups = []

src/crewai/agents/parser.py

@@ -94,6 +94,13 @@ class CrewAgentParser:
 
         elif includes_answer:
             final_answer = text.split(FINAL_ANSWER_ACTION)[-1].strip()
+            # Check whether the final answer ends with triple backticks.
+            if final_answer.endswith("```"):
+                # Count occurrences of triple backticks in the final answer.
+                count = final_answer.count("```")
+                # If count is odd then it's an unmatched trailing set; remove it.
+                if count % 2 != 0:
+                    final_answer = final_answer[:-3].rstrip()
             return AgentFinish(thought, final_answer, text)
 
         if not re.search(r"Action\s*\d*\s*:[\s]*(.*?)", text, re.DOTALL):
@@ -120,7 +127,10 @@ class CrewAgentParser:
         regex = r"(.*?)(?:\n\nAction|\n\nFinal Answer)"
         thought_match = re.search(regex, text, re.DOTALL)
         if thought_match:
-            return thought_match.group(1).strip()
+            thought = thought_match.group(1).strip()
+            # Remove any triple backticks from the thought string
+            thought = thought.replace("```", "").strip()
+            return thought
         return ""
 
     def _clean_action(self, text: str) -> str:

src/crewai/crew.py

@@ -38,11 +38,24 @@ from crewai.tasks.task_output import TaskOutput
 from crewai.telemetry import Telemetry
 from crewai.tools.agent_tools.agent_tools import AgentTools
 from crewai.tools.base_tool import Tool
+from crewai.traces.unified_trace_controller import init_crew_main_trace
 from crewai.types.usage_metrics import UsageMetrics
 from crewai.utilities import I18N, FileHandler, Logger, RPMController
 from crewai.utilities.constants import TRAINING_DATA_FILE
 from crewai.utilities.evaluators.crew_evaluator_handler import CrewEvaluator
 from crewai.utilities.evaluators.task_evaluator import TaskEvaluator
+from crewai.utilities.events.crew_events import (
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewKickoffStartedEvent,
+    CrewTestCompletedEvent,
+    CrewTestFailedEvent,
+    CrewTestStartedEvent,
+    CrewTrainCompletedEvent,
+    CrewTrainFailedEvent,
+    CrewTrainStartedEvent,
+)
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
 from crewai.utilities.formatter import (
     aggregate_raw_outputs_from_task_outputs,
     aggregate_raw_outputs_from_tasks,
@@ -52,12 +65,6 @@ from crewai.utilities.planning_handler import CrewPlanner
 from crewai.utilities.task_output_storage_handler import TaskOutputStorageHandler
 from crewai.utilities.training_handler import CrewTrainingHandler
 
-try:
-    import agentops  # type: ignore
-except ImportError:
-    agentops = None
-
-
 warnings.filterwarnings("ignore", category=SyntaxWarning, module="pysbd")
 
 
@@ -275,12 +282,26 @@ class Crew(BaseModel):
                 if self.entity_memory
                 else EntityMemory(crew=self, embedder_config=self.embedder)
             )
-            if hasattr(self, "memory_config") and self.memory_config is not None:
-                self._user_memory = (
-                    self.user_memory if self.user_memory else UserMemory(crew=self)
-                )
+            if (
+                self.memory_config and "user_memory" in self.memory_config
+            ):  # Check for user_memory in config
+                user_memory_config = self.memory_config["user_memory"]
+                if isinstance(
+                    user_memory_config, UserMemory
+                ):  # Check if it is already an instance
+                    self._user_memory = user_memory_config
+                elif isinstance(
+                    user_memory_config, dict
+                ):  # Check if it's a configuration dict
+                    self._user_memory = UserMemory(
+                        crew=self, **user_memory_config
+                    )  # Initialize with config
+                else:
+                    raise TypeError(
+                        "user_memory must be a UserMemory instance or a configuration dictionary"
+                    )
             else:
-                self._user_memory = None
+                self._user_memory = None  # No user memory if not in config
         return self
 
     @model_validator(mode="after")
@@ -455,8 +476,6 @@ class Crew(BaseModel):
                         )
         return self
 
-
-
     @property
     def key(self) -> str:
         source = [agent.key for agent in self.agents] + [
@@ -509,10 +528,19 @@ class Crew(BaseModel):
         self, n_iterations: int, filename: str, inputs: Optional[Dict[str, Any]] = {}
     ) -> None:
         """Trains the crew for a given number of iterations."""
-        train_crew = self.copy()
-        train_crew._setup_for_training(filename)
-
         try:
+            crewai_event_bus.emit(
+                self,
+                CrewTrainStartedEvent(
+                    crew_name=self.name or "crew",
+                    n_iterations=n_iterations,
+                    filename=filename,
+                    inputs=inputs,
+                ),
+            )
+            train_crew = self.copy()
+            train_crew._setup_for_training(filename)
+
             for n_iteration in range(n_iterations):
                 train_crew._train_iteration = n_iteration
                 train_crew.kickoff(inputs=inputs)
@@ -527,70 +555,94 @@ class Crew(BaseModel):
                     CrewTrainingHandler(filename).save_trained_data(
                         agent_id=str(agent.role), trained_data=result.model_dump()
                     )
+
+            crewai_event_bus.emit(
+                self,
+                CrewTrainCompletedEvent(
+                    crew_name=self.name or "crew",
+                    n_iterations=n_iterations,
+                    filename=filename,
+                ),
+            )
         except Exception as e:
+            crewai_event_bus.emit(
+                self,
+                CrewTrainFailedEvent(error=str(e), crew_name=self.name or "crew"),
+            )
             self._logger.log("error", f"Training failed: {e}", color="red")
             CrewTrainingHandler(TRAINING_DATA_FILE).clear()
             CrewTrainingHandler(filename).clear()
             raise
 
+    @init_crew_main_trace
     def kickoff(
         self,
         inputs: Optional[Dict[str, Any]] = None,
     ) -> CrewOutput:
-        for before_callback in self.before_kickoff_callbacks:
-            if inputs is None:
-                inputs = {}
-            inputs = before_callback(inputs)
+        try:
+            for before_callback in self.before_kickoff_callbacks:
+                if inputs is None:
+                    inputs = {}
+                inputs = before_callback(inputs)
 
-        """Starts the crew to work on its assigned tasks."""
-        self._execution_span = self._telemetry.crew_execution_span(self, inputs)
-        self._task_output_handler.reset()
-        self._logging_color = "bold_purple"
-
-        if inputs is not None:
-            self._inputs = inputs
-            self._interpolate_inputs(inputs)
-        self._set_tasks_callbacks()
-
-        i18n = I18N(prompt_file=self.prompt_file)
-
-        for agent in self.agents:
-            agent.i18n = i18n
-            # type: ignore[attr-defined] # Argument 1 to "_interpolate_inputs" of "Crew" has incompatible type "dict[str, Any] | None"; expected "dict[str, Any]"
-            agent.crew = self  # type: ignore[attr-defined]
-            # TODO: Create an AgentFunctionCalling protocol for future refactoring
-            if not agent.function_calling_llm:  # type: ignore # "BaseAgent" has no attribute "function_calling_llm"
-                agent.function_calling_llm = self.function_calling_llm  # type: ignore # "BaseAgent" has no attribute "function_calling_llm"
-
-            if not agent.step_callback:  # type: ignore # "BaseAgent" has no attribute "step_callback"
-                agent.step_callback = self.step_callback  # type: ignore # "BaseAgent" has no attribute "step_callback"
-
-            agent.create_agent_executor()
-
-        if self.planning:
-            self._handle_crew_planning()
-
-        metrics: List[UsageMetrics] = []
-
-        if self.process == Process.sequential:
-            result = self._run_sequential_process()
-        elif self.process == Process.hierarchical:
-            result = self._run_hierarchical_process()
-        else:
-            raise NotImplementedError(
-                f"The process '{self.process}' is not implemented yet."
+            crewai_event_bus.emit(
+                self,
+                CrewKickoffStartedEvent(crew_name=self.name or "crew", inputs=inputs),
             )
 
-        for after_callback in self.after_kickoff_callbacks:
-            result = after_callback(result)
+            # Starts the crew to work on its assigned tasks.
+            self._task_output_handler.reset()
+            self._logging_color = "bold_purple"
 
-        metrics += [agent._token_process.get_summary() for agent in self.agents]
+            if inputs is not None:
+                self._inputs = inputs
+                self._interpolate_inputs(inputs)
+            self._set_tasks_callbacks()
 
-        self.usage_metrics = UsageMetrics()
-        for metric in metrics:
-            self.usage_metrics.add_usage_metrics(metric)
+            i18n = I18N(prompt_file=self.prompt_file)
 
-        return result
+            for agent in self.agents:
+                agent.i18n = i18n
+                # type: ignore[attr-defined] # Argument 1 to "_interpolate_inputs" of "Crew" has incompatible type "dict[str, Any] | None"; expected "dict[str, Any]"
+                agent.crew = self  # type: ignore[attr-defined]
+                # TODO: Create an AgentFunctionCalling protocol for future refactoring
+                if not agent.function_calling_llm:  # type: ignore # "BaseAgent" has no attribute "function_calling_llm"
+                    agent.function_calling_llm = self.function_calling_llm  # type: ignore # "BaseAgent" has no attribute "function_calling_llm"
+
+                if not agent.step_callback:  # type: ignore # "BaseAgent" has no attribute "step_callback"
+                    agent.step_callback = self.step_callback  # type: ignore # "BaseAgent" has no attribute "step_callback"
+
+                agent.create_agent_executor()
+
+            if self.planning:
+                self._handle_crew_planning()
+
+            metrics: List[UsageMetrics] = []
+
+            if self.process == Process.sequential:
+                result = self._run_sequential_process()
+            elif self.process == Process.hierarchical:
+                result = self._run_hierarchical_process()
+            else:
+                raise NotImplementedError(
+                    f"The process '{self.process}' is not implemented yet."
+                )
+
+            for after_callback in self.after_kickoff_callbacks:
+                result = after_callback(result)
+
+            metrics += [agent._token_process.get_summary() for agent in self.agents]
+
+            self.usage_metrics = UsageMetrics()
+            for metric in metrics:
+                self.usage_metrics.add_usage_metrics(metric)
+            return result
+        except Exception as e:
+            crewai_event_bus.emit(
+                self,
+                CrewKickoffFailedEvent(error=str(e), crew_name=self.name or "crew"),
+            )
+            raise
 
     def kickoff_for_each(self, inputs: List[Dict[str, Any]]) -> List[CrewOutput]:
         """Executes the Crew's workflow for each input in the list and aggregates results."""
@@ -928,17 +980,22 @@ class Crew(BaseModel):
     def _create_crew_output(self, task_outputs: List[TaskOutput]) -> CrewOutput:
         if not task_outputs:
             raise ValueError("No task outputs available to create crew output.")
-            
+
         # Filter out empty outputs and get the last valid one as the main output
         valid_outputs = [t for t in task_outputs if t.raw]
         if not valid_outputs:
             raise ValueError("No valid task outputs available to create crew output.")
         final_task_output = valid_outputs[-1]
-            
+
         final_string_output = final_task_output.raw
         self._finish_execution(final_string_output)
         token_usage = self.calculate_usage_metrics()
-
+        crewai_event_bus.emit(
+            self,
+            CrewKickoffCompletedEvent(
+                crew_name=self.name or "crew", output=final_task_output
+            ),
+        )
         return CrewOutput(
             raw=final_task_output.raw,
             pydantic=final_task_output.pydantic,
@@ -1124,13 +1181,6 @@ class Crew(BaseModel):
     def _finish_execution(self, final_string_output: str) -> None:
         if self.max_rpm:
             self._rpm_controller.stop_rpm_counter()
-        if agentops:
-            agentops.end_session(
-                end_state="Success",
-                end_state_reason="Finished Execution",
-                is_auto_end=True,
-            )
-        self._telemetry.end_crew(self, final_string_output)
 
     def calculate_usage_metrics(self) -> UsageMetrics:
         """Calculates and returns the usage metrics."""
@@ -1148,25 +1198,45 @@ class Crew(BaseModel):
     def test(
         self,
         n_iterations: int,
-        openai_model_name: Optional[str] = None,
+        eval_llm: Union[str, InstanceOf[LLM]],
         inputs: Optional[Dict[str, Any]] = None,
     ) -> None:
         """Test and evaluate the Crew with the given inputs for n iterations concurrently using concurrent.futures."""
-        test_crew = self.copy()
+        try:
+            eval_llm = create_llm(eval_llm)
+            if not eval_llm:
+                raise ValueError("Failed to create LLM instance.")
 
-        self._test_execution_span = test_crew._telemetry.test_execution_span(
-            test_crew,
-            n_iterations,
-            inputs,
-            openai_model_name,  # type: ignore[arg-type]
-        )  # type: ignore[arg-type]
-        evaluator = CrewEvaluator(test_crew, openai_model_name)  # type: ignore[arg-type]
+            crewai_event_bus.emit(
+                self,
+                CrewTestStartedEvent(
+                    crew_name=self.name or "crew",
+                    n_iterations=n_iterations,
+                    eval_llm=eval_llm,
+                    inputs=inputs,
+                ),
+            )
+            test_crew = self.copy()
+            evaluator = CrewEvaluator(test_crew, eval_llm)  # type: ignore[arg-type]
 
-        for i in range(1, n_iterations + 1):
-            evaluator.set_iteration(i)
-            test_crew.kickoff(inputs=inputs)
+            for i in range(1, n_iterations + 1):
+                evaluator.set_iteration(i)
+                test_crew.kickoff(inputs=inputs)
 
-        evaluator.print_crew_evaluation_result()
+            evaluator.print_crew_evaluation_result()
+
+            crewai_event_bus.emit(
+                self,
+                CrewTestCompletedEvent(
+                    crew_name=self.name or "crew",
+                ),
+            )
+        except Exception as e:
+            crewai_event_bus.emit(
+                self,
+                CrewTestFailedEvent(error=str(e), crew_name=self.name or "crew"),
+            )
+            raise
 
     def __repr__(self):
         return f"Crew(id={self.id}, process={self.process}, number_of_agents={len(self.agents)}, number_of_tasks={len(self.tasks)})"

src/crewai/flow/flow.py

@@ -17,19 +17,25 @@ from typing import (
 )
 from uuid import uuid4
 
-from blinker import Signal
 from pydantic import BaseModel, Field, ValidationError
 
-from crewai.flow.flow_events import (
-    FlowFinishedEvent,
-    FlowStartedEvent,
-    MethodExecutionFinishedEvent,
-    MethodExecutionStartedEvent,
-)
 from crewai.flow.flow_visualizer import plot_flow
 from crewai.flow.persistence.base import FlowPersistence
 from crewai.flow.utils import get_possible_return_constants
-from crewai.telemetry import Telemetry
+from crewai.traces.unified_trace_controller import (
+    init_flow_main_trace,
+    trace_flow_step,
+)
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
+from crewai.utilities.events.flow_events import (
+    FlowCreatedEvent,
+    FlowFinishedEvent,
+    FlowPlotEvent,
+    FlowStartedEvent,
+    MethodExecutionFailedEvent,
+    MethodExecutionFinishedEvent,
+    MethodExecutionStartedEvent,
+)
 from crewai.utilities.printer import Printer
 
 logger = logging.getLogger(__name__)
@@ -427,7 +433,6 @@ class Flow(Generic[T], metaclass=FlowMeta):
 
     Type parameter T must be either Dict[str, Any] or a subclass of BaseModel."""
 
-    _telemetry = Telemetry()
     _printer = Printer()
 
     _start_methods: List[str] = []
@@ -435,7 +440,6 @@ class Flow(Generic[T], metaclass=FlowMeta):
     _routers: Set[str] = set()
     _router_paths: Dict[str, List[str]] = {}
     initial_state: Union[Type[T], T, None] = None
-    event_emitter = Signal("event_emitter")
 
     def __class_getitem__(cls: Type["Flow"], item: Type[T]) -> Type["Flow"]:
         class _FlowGeneric(cls):  # type: ignore
@@ -469,7 +473,13 @@ class Flow(Generic[T], metaclass=FlowMeta):
         if kwargs:
             self._initialize_state(kwargs)
 
-        self._telemetry.flow_creation_span(self.__class__.__name__)
+        crewai_event_bus.emit(
+            self,
+            FlowCreatedEvent(
+                type="flow_created",
+                flow_name=self.__class__.__name__,
+            ),
+        )
 
         # Register all flow-related methods
         for method_name in dir(self):
@@ -738,9 +748,9 @@ class Flow(Generic[T], metaclass=FlowMeta):
                 self._initialize_state(filtered_inputs)
 
         # Start flow execution
-        self.event_emitter.send(
+        crewai_event_bus.emit(
             self,
-            event=FlowStartedEvent(
+            FlowStartedEvent(
                 type="flow_started",
                 flow_name=self.__class__.__name__,
                 inputs=inputs,
@@ -753,16 +763,16 @@ class Flow(Generic[T], metaclass=FlowMeta):
         if inputs is not None and "id" not in inputs:
             self._initialize_state(inputs)
 
-        return asyncio.run(self.kickoff_async())
+        async def run_flow():
+            return await self.kickoff_async()
 
+        return asyncio.run(run_flow())
+
+    @init_flow_main_trace
     async def kickoff_async(self, inputs: Optional[Dict[str, Any]] = None) -> Any:
         if not self._start_methods:
             raise ValueError("No start method defined")
 
-        self._telemetry.flow_execution_span(
-            self.__class__.__name__, list(self._methods.keys())
-        )
-
         tasks = [
             self._execute_start_method(start_method)
             for start_method in self._start_methods
@@ -771,9 +781,9 @@ class Flow(Generic[T], metaclass=FlowMeta):
 
         final_output = self._method_outputs[-1] if self._method_outputs else None
 
-        self.event_emitter.send(
+        crewai_event_bus.emit(
             self,
-            event=FlowFinishedEvent(
+            FlowFinishedEvent(
                 type="flow_finished",
                 flow_name=self.__class__.__name__,
                 result=final_output,
@@ -804,43 +814,59 @@ class Flow(Generic[T], metaclass=FlowMeta):
         )
         await self._execute_listeners(start_method_name, result)
 
+    @trace_flow_step
     async def _execute_method(
         self, method_name: str, method: Callable, *args: Any, **kwargs: Any
     ) -> Any:
-        dumped_params = {f"_{i}": arg for i, arg in enumerate(args)} | (kwargs or {})
-        self.event_emitter.send(
-            self,
-            event=MethodExecutionStartedEvent(
-                type="method_execution_started",
-                method_name=method_name,
-                flow_name=self.__class__.__name__,
-                params=dumped_params,
-                state=self._copy_state(),
-            ),
-        )
+        try:
+            dumped_params = {f"_{i}": arg for i, arg in enumerate(args)} | (
+                kwargs or {}
+            )
+            crewai_event_bus.emit(
+                self,
+                MethodExecutionStartedEvent(
+                    type="method_execution_started",
+                    method_name=method_name,
+                    flow_name=self.__class__.__name__,
+                    params=dumped_params,
+                    state=self._copy_state(),
+                ),
+            )
 
-        result = (
-            await method(*args, **kwargs)
-            if asyncio.iscoroutinefunction(method)
-            else method(*args, **kwargs)
-        )
-        self._method_outputs.append(result)
-        self._method_execution_counts[method_name] = (
-            self._method_execution_counts.get(method_name, 0) + 1
-        )
+            result = (
+                await method(*args, **kwargs)
+                if asyncio.iscoroutinefunction(method)
+                else method(*args, **kwargs)
+            )
 
-        self.event_emitter.send(
-            self,
-            event=MethodExecutionFinishedEvent(
-                type="method_execution_finished",
-                method_name=method_name,
-                flow_name=self.__class__.__name__,
-                state=self._copy_state(),
-                result=result,
-            ),
-        )
+            self._method_outputs.append(result)
+            self._method_execution_counts[method_name] = (
+                self._method_execution_counts.get(method_name, 0) + 1
+            )
 
-        return result
+            crewai_event_bus.emit(
+                self,
+                MethodExecutionFinishedEvent(
+                    type="method_execution_finished",
+                    method_name=method_name,
+                    flow_name=self.__class__.__name__,
+                    state=self._copy_state(),
+                    result=result,
+                ),
+            )
+
+            return result
+        except Exception as e:
+            crewai_event_bus.emit(
+                self,
+                MethodExecutionFailedEvent(
+                    type="method_execution_failed",
+                    method_name=method_name,
+                    flow_name=self.__class__.__name__,
+                    error=e,
+                ),
+            )
+            raise e
 
     async def _execute_listeners(self, trigger_method: str, result: Any) -> None:
         """
@@ -978,6 +1004,7 @@ class Flow(Generic[T], metaclass=FlowMeta):
         """
         try:
             method = self._methods[listener_name]
+
             sig = inspect.signature(method)
             params = list(sig.parameters.values())
             method_params = [p for p in params if p.name != "self"]
@@ -1027,7 +1054,11 @@ class Flow(Generic[T], metaclass=FlowMeta):
             logger.warning(message)
 
     def plot(self, filename: str = "crewai_flow") -> None:
-        self._telemetry.flow_plotting_span(
-            self.__class__.__name__, list(self._methods.keys())
+        crewai_event_bus.emit(
+            self,
+            FlowPlotEvent(
+                type="flow_plot",
+                flow_name=self.__class__.__name__,
+            ),
         )
         plot_flow(self, filename)

src/crewai/flow/flow_events.py

@@ -1,39 +0,0 @@
-from dataclasses import dataclass, field
-from datetime import datetime
-from typing import Any, Dict, Optional, Union
-
-from pydantic import BaseModel
-
-
-@dataclass
-class Event:
-    type: str
-    flow_name: str
-    timestamp: datetime = field(init=False)
-
-    def __post_init__(self):
-        self.timestamp = datetime.now()
-
-
-@dataclass
-class FlowStartedEvent(Event):
-    inputs: Optional[Dict[str, Any]] = None
-
-
-@dataclass
-class MethodExecutionStartedEvent(Event):
-    method_name: str
-    state: Union[Dict[str, Any], BaseModel]
-    params: Optional[Dict[str, Any]] = None
-
-
-@dataclass
-class MethodExecutionFinishedEvent(Event):
-    method_name: str
-    state: Union[Dict[str, Any], BaseModel]
-    result: Any = None
-
-
-@dataclass
-class FlowFinishedEvent(Event):
-    result: Optional[Any] = None

src/crewai/flow/persistence/decorators.py

@@ -58,7 +58,7 @@ class PersistenceDecorator:
     _printer = Printer()  # Class-level printer instance
 
     @classmethod
-    def persist_state(cls, flow_instance: Any, method_name: str, persistence_instance: FlowPersistence) -> None:
+    def persist_state(cls, flow_instance: Any, method_name: str, persistence_instance: FlowPersistence, verbose: bool = False) -> None:
         """Persist flow state with proper error handling and logging.
 
         This method handles the persistence of flow state data, including proper
@@ -68,6 +68,7 @@ class PersistenceDecorator:
             flow_instance: The flow instance whose state to persist
             method_name: Name of the method that triggered persistence
             persistence_instance: The persistence backend to use
+            verbose: Whether to log persistence operations
 
         Raises:
             ValueError: If flow has no state or state lacks an ID
@@ -88,9 +89,10 @@ class PersistenceDecorator:
             if not flow_uuid:
                 raise ValueError("Flow state must have an 'id' field for persistence")
 
-            # Log state saving with consistent message
-            cls._printer.print(LOG_MESSAGES["save_state"].format(flow_uuid), color="cyan")
-            logger.info(LOG_MESSAGES["save_state"].format(flow_uuid))
+            # Log state saving only if verbose is True
+            if verbose:
+                cls._printer.print(LOG_MESSAGES["save_state"].format(flow_uuid), color="cyan")
+                logger.info(LOG_MESSAGES["save_state"].format(flow_uuid))
 
             try:
                 persistence_instance.save_state(
@@ -115,7 +117,7 @@ class PersistenceDecorator:
             raise ValueError(error_msg) from e
 
 
-def persist(persistence: Optional[FlowPersistence] = None):
+def persist(persistence: Optional[FlowPersistence] = None, verbose: bool = False):
     """Decorator to persist flow state.
 
     This decorator can be applied at either the class level or method level.
@@ -126,6 +128,7 @@ def persist(persistence: Optional[FlowPersistence] = None):
     Args:
         persistence: Optional FlowPersistence implementation to use.
                     If not provided, uses SQLiteFlowPersistence.
+        verbose: Whether to log persistence operations. Defaults to False.
 
     Returns:
         A decorator that can be applied to either a class or method
@@ -135,13 +138,12 @@ def persist(persistence: Optional[FlowPersistence] = None):
         RuntimeError: If state persistence fails
 
     Example:
-        @persist  # Class-level persistence with default SQLite
+        @persist(verbose=True)  # Class-level persistence with logging
         class MyFlow(Flow[MyState]):
             @start()
             def begin(self):
                 pass
     """
-
     def decorator(target: Union[Type, Callable[..., T]]) -> Union[Type, Callable[..., T]]:
         """Decorator that handles both class and method decoration."""
         actual_persistence = persistence or SQLiteFlowPersistence()
@@ -179,7 +181,7 @@ def persist(persistence: Optional[FlowPersistence] = None):
                         @functools.wraps(original_method)
                         async def method_wrapper(self: Any, *args: Any, **kwargs: Any) -> Any:
                             result = await original_method(self, *args, **kwargs)
-                            PersistenceDecorator.persist_state(self, method_name, actual_persistence)
+                            PersistenceDecorator.persist_state(self, method_name, actual_persistence, verbose)
                             return result
                         return method_wrapper
 
@@ -199,7 +201,7 @@ def persist(persistence: Optional[FlowPersistence] = None):
                         @functools.wraps(original_method)
                         def method_wrapper(self: Any, *args: Any, **kwargs: Any) -> Any:
                             result = original_method(self, *args, **kwargs)
-                            PersistenceDecorator.persist_state(self, method_name, actual_persistence)
+                            PersistenceDecorator.persist_state(self, method_name, actual_persistence, verbose)
                             return result
                         return method_wrapper
 
@@ -228,7 +230,7 @@ def persist(persistence: Optional[FlowPersistence] = None):
                         result = await method_coro
                     else:
                         result = method_coro
-                    PersistenceDecorator.persist_state(flow_instance, method.__name__, actual_persistence)
+                    PersistenceDecorator.persist_state(flow_instance, method.__name__, actual_persistence, verbose)
                     return result
 
                 for attr in ["__is_start_method__", "__trigger_methods__", "__condition_type__", "__is_router__"]:
@@ -240,7 +242,7 @@ def persist(persistence: Optional[FlowPersistence] = None):
                 @functools.wraps(method)
                 def method_sync_wrapper(flow_instance: Any, *args: Any, **kwargs: Any) -> T:
                     result = method(flow_instance, *args, **kwargs)
-                    PersistenceDecorator.persist_state(flow_instance, method.__name__, actual_persistence)
+                    PersistenceDecorator.persist_state(flow_instance, method.__name__, actual_persistence, verbose)
                     return result
 
                 for attr in ["__is_start_method__", "__trigger_methods__", "__condition_type__", "__is_router__"]:

src/crewai/flow/persistence/sqlite.py

@@ -11,6 +11,7 @@ from typing import Any, Dict, Optional, Union
 from pydantic import BaseModel
 
 from crewai.flow.persistence.base import FlowPersistence
+from crewai.utilities.datetime_compat import UTC
 
 
 class SQLiteFlowPersistence(FlowPersistence):
@@ -95,7 +96,7 @@ class SQLiteFlowPersistence(FlowPersistence):
             """, (
                 flow_uuid,
                 method_name,
-                datetime.utcnow().isoformat(),
+                datetime.now(UTC).isoformat(),
                 json.dumps(state_dict),
             ))
 

src/crewai/flow/state_utils.py

@@ -0,0 +1,91 @@
+import json
+from datetime import date, datetime
+from typing import Any, Dict, List, Union
+
+from pydantic import BaseModel
+
+from crewai.flow import Flow
+
+SerializablePrimitive = Union[str, int, float, bool, None]
+Serializable = Union[
+    SerializablePrimitive, List["Serializable"], Dict[str, "Serializable"]
+]
+
+
+def export_state(flow: Flow) -> dict[str, Serializable]:
+    """Exports the Flow's internal state as JSON-compatible data structures.
+
+    Performs a one-way transformation of a Flow's state into basic Python types
+    that can be safely serialized to JSON. To prevent infinite recursion with
+    circular references, the conversion is limited to a depth of 5 levels.
+
+    Args:
+        flow: The Flow object whose state needs to be exported
+
+    Returns:
+        dict[str, Any]: The transformed state using JSON-compatible Python
+            types.
+    """
+    result = to_serializable(flow._state)
+    assert isinstance(result, dict)
+    return result
+
+
+def to_serializable(
+    obj: Any, max_depth: int = 5, _current_depth: int = 0
+) -> Serializable:
+    """Converts a Python object into a JSON-compatible representation.
+
+    Supports primitives, datetime objects, collections, dictionaries, and
+    Pydantic models. Recursion depth is limited to prevent infinite nesting.
+    Non-convertible objects default to their string representations.
+
+    Args:
+        obj (Any): Object to transform.
+        max_depth (int, optional): Maximum recursion depth. Defaults to 5.
+
+    Returns:
+        Serializable: A JSON-compatible structure.
+    """
+    if _current_depth >= max_depth:
+        return repr(obj)
+
+    if isinstance(obj, (str, int, float, bool, type(None))):
+        return obj
+    elif isinstance(obj, (date, datetime)):
+        return obj.isoformat()
+    elif isinstance(obj, (list, tuple, set)):
+        return [to_serializable(item, max_depth, _current_depth + 1) for item in obj]
+    elif isinstance(obj, dict):
+        return {
+            _to_serializable_key(key): to_serializable(
+                value, max_depth, _current_depth + 1
+            )
+            for key, value in obj.items()
+        }
+    elif isinstance(obj, BaseModel):
+        return to_serializable(obj.model_dump(), max_depth, _current_depth + 1)
+    else:
+        return repr(obj)
+
+
+def _to_serializable_key(key: Any) -> str:
+    if isinstance(key, (str, int)):
+        return str(key)
+    return f"key_{id(key)}_{repr(key)}"
+
+
+def to_string(obj: Any) -> str | None:
+    """Serializes an object into a JSON string.
+
+    Args:
+        obj (Any): Object to serialize.
+
+    Returns:
+        str | None: A JSON-formatted string or `None` if empty.
+    """
+    serializable = to_serializable(obj)
+    if serializable is None:
+        return None
+    else:
+        return json.dumps(serializable)

src/crewai/knowledge/storage/knowledge_storage.py

@@ -76,7 +76,7 @@ class KnowledgeStorage(BaseKnowledgeStorage):
                         "context": fetched["documents"][0][i],  # type: ignore
                         "score": fetched["distances"][0][i],  # type: ignore
                     }
-                    if result["score"] >= score_threshold:  # type: ignore
+                    if result["score"] >= score_threshold:
                         results.append(result)
                 return results
             else:

src/crewai/llm.py

@@ -1,3 +1,4 @@
+import inspect
 import json
 import logging
 import os
@@ -5,11 +6,23 @@ import sys
 import threading
 import warnings
 from contextlib import contextmanager
-from typing import Any, Dict, List, Literal, Optional, Type, Union, cast
+from typing import (
+    Any,
+    Dict,
+    List,
+    Literal,
+    Optional,
+    Tuple,
+    Type,
+    Union,
+    cast,
+)
 
 from dotenv import load_dotenv
 from pydantic import BaseModel
 
+from crewai.utilities.events.tool_usage_events import ToolExecutionErrorEvent
+
 with warnings.catch_warnings():
     warnings.simplefilter("ignore", UserWarning)
     import litellm
@@ -18,9 +31,12 @@ with warnings.catch_warnings():
     from litellm.utils import supports_response_schema
 
 
+from crewai.traces.unified_trace_controller import trace_llm_call
+from crewai.utilities.events import crewai_event_bus
 from crewai.utilities.exceptions.context_window_exceeding_exception import (
     LLMContextLengthExceededException,
 )
+from crewai.utilities.protocols import AgentExecutorProtocol
 
 load_dotenv()
 
@@ -164,6 +180,7 @@ class LLM:
         self.context_window_size = 0
         self.reasoning_effort = reasoning_effort
         self.additional_params = kwargs
+        self._message_history: List[Dict[str, str]] = []
         self.is_anthropic = self._is_anthropic_model(model)
 
         litellm.drop_params = True
@@ -179,16 +196,22 @@ class LLM:
         self.set_callbacks(callbacks)
         self.set_env_callbacks()
 
+    @trace_llm_call
+    def _call_llm(self, params: Dict[str, Any]) -> Any:
+        with suppress_warnings():
+            response = litellm.completion(**params)
+            return response
+
     def _is_anthropic_model(self, model: str) -> bool:
         """Determine if the model is from Anthropic provider.
-        
+
         Args:
             model: The model identifier string.
-            
+
         Returns:
             bool: True if the model is from Anthropic, False otherwise.
         """
-        ANTHROPIC_PREFIXES = ('anthropic/', 'claude-', 'claude/')
+        ANTHROPIC_PREFIXES = ("anthropic/", "claude-", "claude/")
         return any(prefix in model.lower() for prefix in ANTHROPIC_PREFIXES)
 
     def call(
@@ -199,7 +222,7 @@ class LLM:
         available_functions: Optional[Dict[str, Any]] = None,
     ) -> Union[str, Any]:
         """High-level LLM call method.
-        
+
         Args:
             messages: Input messages for the LLM.
                      Can be a string or list of message dictionaries.
@@ -211,22 +234,22 @@ class LLM:
                       during and after the LLM call.
             available_functions: Optional dict mapping function names to callables
                                that can be invoked by the LLM.
-        
+
         Returns:
             Union[str, Any]: Either a text response from the LLM (str) or
                            the result of a tool function call (Any).
-        
+
         Raises:
             TypeError: If messages format is invalid
             ValueError: If response format is not supported
             LLMContextLengthExceededException: If input exceeds model's context limit
-        
+
         Examples:
             # Example 1: Simple string input
             >>> response = llm.call("Return the name of a random city.")
             >>> print(response)
             "Paris"
-            
+
             # Example 2: Message list with system and user messages
             >>> messages = [
             ...     {"role": "system", "content": "You are a geography expert"},
@@ -288,7 +311,7 @@ class LLM:
                 params = {k: v for k, v in params.items() if v is not None}
 
                 # --- 2) Make the completion call
-                response = litellm.completion(**params)
+                response = self._call_llm(params)
                 response_message = cast(Choices, cast(ModelResponse, response).choices)[
                     0
                 ].message
@@ -315,7 +338,7 @@ class LLM:
                 # --- 5) Handle the tool call
                 tool_call = tool_calls[0]
                 function_name = tool_call.function.name
-
+                print("function_name", function_name)
                 if function_name in available_functions:
                     try:
                         function_args = json.loads(tool_call.function.arguments)
@@ -333,6 +356,15 @@ class LLM:
                         logging.error(
                             f"Error executing function '{function_name}': {e}"
                         )
+                        crewai_event_bus.emit(
+                            self,
+                            event=ToolExecutionErrorEvent(
+                                tool_name=function_name,
+                                tool_args=function_args,
+                                tool_class=fn,
+                                error=str(e),
+                            ),
+                        )
                         return text_response
 
                 else:
@@ -348,36 +380,40 @@ class LLM:
                     logging.error(f"LiteLLM call failed: {str(e)}")
                 raise
 
-    def _format_messages_for_provider(self, messages: List[Dict[str, str]]) -> List[Dict[str, str]]:
+    def _format_messages_for_provider(
+        self, messages: List[Dict[str, str]]
+    ) -> List[Dict[str, str]]:
         """Format messages according to provider requirements.
-        
+
         Args:
             messages: List of message dictionaries with 'role' and 'content' keys.
                      Can be empty or None.
-        
+
         Returns:
             List of formatted messages according to provider requirements.
             For Anthropic models, ensures first message has 'user' role.
-        
+
         Raises:
             TypeError: If messages is None or contains invalid message format.
         """
         if messages is None:
             raise TypeError("Messages cannot be None")
-            
+
         # Validate message format first
         for msg in messages:
             if not isinstance(msg, dict) or "role" not in msg or "content" not in msg:
-                raise TypeError("Invalid message format. Each message must be a dict with 'role' and 'content' keys")
-            
+                raise TypeError(
+                    "Invalid message format. Each message must be a dict with 'role' and 'content' keys"
+                )
+
         if not self.is_anthropic:
             return messages
-            
+
         # Anthropic requires messages to start with 'user' role
         if not messages or messages[0]["role"] == "system":
             # If first message is system or empty, add a placeholder user message
             return [{"role": "user", "content": "."}, *messages]
-                
+
         return messages
 
     def _get_custom_llm_provider(self) -> str:
@@ -495,3 +531,95 @@ class LLM:
 
                 litellm.success_callback = success_callbacks
                 litellm.failure_callback = failure_callbacks
+
+    def _get_execution_context(self) -> Tuple[Optional[Any], Optional[Any]]:
+        """Get the agent and task from the execution context.
+
+        Returns:
+            tuple: (agent, task) from any AgentExecutor context, or (None, None) if not found
+        """
+        frame = inspect.currentframe()
+        caller_frame = frame.f_back if frame else None
+        agent = None
+        task = None
+
+        # Add a maximum depth to prevent infinite loops
+        max_depth = 100  # Reasonable limit for call stack depth
+        current_depth = 0
+
+        while caller_frame and current_depth < max_depth:
+            if "self" in caller_frame.f_locals:
+                caller_self = caller_frame.f_locals["self"]
+                if isinstance(caller_self, AgentExecutorProtocol):
+                    agent = caller_self.agent
+                    task = caller_self.task
+                    break
+            caller_frame = caller_frame.f_back
+            current_depth += 1
+
+        return agent, task
+
+    def _get_new_messages(self, messages: List[Dict[str, str]]) -> List[Dict[str, str]]:
+        """Get only the new messages that haven't been processed before."""
+        if not hasattr(self, "_message_history"):
+            self._message_history = []
+
+        new_messages = []
+        for message in messages:
+            message_key = (message["role"], message["content"])
+            if message_key not in [
+                (m["role"], m["content"]) for m in self._message_history
+            ]:
+                new_messages.append(message)
+                self._message_history.append(message)
+        return new_messages
+
+    def _get_new_tool_results(self, agent) -> List[Dict]:
+        """Get only the new tool results that haven't been processed before."""
+        if not agent or not agent.tools_results:
+            return []
+
+        if not hasattr(self, "_tool_results_history"):
+            self._tool_results_history: List[Dict] = []
+
+        new_tool_results = []
+
+        for result in agent.tools_results:
+            # Process tool arguments to extract actual values
+            processed_args = {}
+            if isinstance(result["tool_args"], dict):
+                for key, value in result["tool_args"].items():
+                    if isinstance(value, dict) and "type" in value:
+                        # Skip metadata and just store the actual value
+                        continue
+                    processed_args[key] = value
+
+            # Create a clean result with processed arguments
+            clean_result = {
+                "tool_name": result["tool_name"],
+                "tool_args": processed_args,
+                "result": result["result"],
+                "content": result.get("content", ""),
+                "start_time": result.get("start_time", ""),
+            }
+
+            # Check if this exact tool execution exists in history
+            is_duplicate = False
+            for history_result in self._tool_results_history:
+                if (
+                    clean_result["tool_name"] == history_result["tool_name"]
+                    and str(clean_result["tool_args"])
+                    == str(history_result["tool_args"])
+                    and str(clean_result["result"]) == str(history_result["result"])
+                    and clean_result["content"] == history_result.get("content", "")
+                    and clean_result["start_time"]
+                    == history_result.get("start_time", "")
+                ):
+                    is_duplicate = True
+                    break
+
+            if not is_duplicate:
+                new_tool_results.append(clean_result)
+                self._tool_results_history.append(clean_result)
+
+        return new_tool_results

src/crewai/task.py

@@ -21,7 +21,6 @@ from typing import (
     Union,
 )
 
-from opentelemetry.trace import Span
 from pydantic import (
     UUID4,
     BaseModel,
@@ -36,10 +35,15 @@ from crewai.agents.agent_builder.base_agent import BaseAgent
 from crewai.tasks.guardrail_result import GuardrailResult
 from crewai.tasks.output_format import OutputFormat
 from crewai.tasks.task_output import TaskOutput
-from crewai.telemetry.telemetry import Telemetry
 from crewai.tools.base_tool import BaseTool
 from crewai.utilities.config import process_config
 from crewai.utilities.converter import Converter, convert_to_model
+from crewai.utilities.events import (
+    TaskCompletedEvent,
+    TaskFailedEvent,
+    TaskStartedEvent,
+)
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
 from crewai.utilities.i18n import I18N
 from crewai.utilities.printer import Printer
 
@@ -183,8 +187,6 @@ class Task(BaseModel):
                     )
         return v
 
-    _telemetry: Telemetry = PrivateAttr(default_factory=Telemetry)
-    _execution_span: Optional[Span] = PrivateAttr(default=None)
     _original_description: Optional[str] = PrivateAttr(default=None)
     _original_expected_output: Optional[str] = PrivateAttr(default=None)
     _original_output_file: Optional[str] = PrivateAttr(default=None)
@@ -348,100 +350,102 @@ class Task(BaseModel):
         tools: Optional[List[Any]],
     ) -> TaskOutput:
         """Run the core execution logic of the task."""
-        agent = agent or self.agent
-        self.agent = agent
-        if not agent:
-            raise Exception(
-                f"The task '{self.description}' has no agent assigned, therefore it can't be executed directly and should be executed in a Crew using a specific process that support that, like hierarchical."
+        try:
+            agent = agent or self.agent
+            self.agent = agent
+            if not agent:
+                raise Exception(
+                    f"The task '{self.description}' has no agent assigned, therefore it can't be executed directly and should be executed in a Crew using a specific process that support that, like hierarchical."
+                )
+
+            self.start_time = datetime.datetime.now()
+
+            self.prompt_context = context
+            tools = tools or self.tools or []
+
+            self.processed_by_agents.add(agent.role)
+            crewai_event_bus.emit(self, TaskStartedEvent(context=context))
+            result = agent.execute_task(
+                task=self,
+                context=context,
+                tools=tools,
             )
 
-        self.start_time = datetime.datetime.now()
-        self._execution_span = self._telemetry.task_started(crew=agent.crew, task=self)
+            pydantic_output, json_output = self._export_output(result)
+            task_output = TaskOutput(
+                name=self.name,
+                description=self.description,
+                expected_output=self.expected_output,
+                raw=result,
+                pydantic=pydantic_output,
+                json_dict=json_output,
+                agent=agent.role,
+                output_format=self._get_output_format(),
+            )
 
-        self.prompt_context = context
-        tools = tools or self.tools or []
+            if self.guardrail:
+                guardrail_result = GuardrailResult.from_tuple(
+                    self.guardrail(task_output)
+                )
+                if not guardrail_result.success:
+                    if self.retry_count >= self.max_retries:
+                        raise Exception(
+                            f"Task failed guardrail validation after {self.max_retries} retries. "
+                            f"Last error: {guardrail_result.error}"
+                        )
 
-        self.processed_by_agents.add(agent.role)
+                    self.retry_count += 1
+                    context = self.i18n.errors("validation_error").format(
+                        guardrail_result_error=guardrail_result.error,
+                        task_output=task_output.raw,
+                    )
+                    printer = Printer()
+                    printer.print(
+                        content=f"Guardrail blocked, retrying, due to: {guardrail_result.error}\n",
+                        color="yellow",
+                    )
+                    return self._execute_core(agent, context, tools)
 
-        result = agent.execute_task(
-            task=self,
-            context=context,
-            tools=tools,
-        )
-
-        pydantic_output, json_output = self._export_output(result)
-        task_output = TaskOutput(
-            name=self.name,
-            description=self.description,
-            expected_output=self.expected_output,
-            raw=result,
-            pydantic=pydantic_output,
-            json_dict=json_output,
-            agent=agent.role,
-            output_format=self._get_output_format(),
-        )
-
-        if self.guardrail:
-            guardrail_result = GuardrailResult.from_tuple(self.guardrail(task_output))
-            if not guardrail_result.success:
-                if self.retry_count >= self.max_retries:
+                if guardrail_result.result is None:
                     raise Exception(
-                        f"Task failed guardrail validation after {self.max_retries} retries. "
-                        f"Last error: {guardrail_result.error}"
+                        "Task guardrail returned None as result. This is not allowed."
                     )
 
-                self.retry_count += 1
-                context = self.i18n.errors("validation_error").format(
-                    guardrail_result_error=guardrail_result.error,
-                    task_output=task_output.raw,
+                if isinstance(guardrail_result.result, str):
+                    task_output.raw = guardrail_result.result
+                    pydantic_output, json_output = self._export_output(
+                        guardrail_result.result
+                    )
+                    task_output.pydantic = pydantic_output
+                    task_output.json_dict = json_output
+                elif isinstance(guardrail_result.result, TaskOutput):
+                    task_output = guardrail_result.result
+
+            self.output = task_output
+            self.end_time = datetime.datetime.now()
+
+            if self.callback:
+                self.callback(self.output)
+
+            crew = self.agent.crew  # type: ignore[union-attr]
+            if crew and crew.task_callback and crew.task_callback != self.callback:
+                crew.task_callback(self.output)
+
+            if self.output_file:
+                content = (
+                    json_output
+                    if json_output
+                    else pydantic_output.model_dump_json()
+                    if pydantic_output
+                    else result
                 )
-                printer = Printer()
-                printer.print(
-                    content=f"Guardrail blocked, retrying, due to: {guardrail_result.error}\n",
-                    color="yellow",
-                )
-                return self._execute_core(agent, context, tools)
-
-            if guardrail_result.result is None:
-                raise Exception(
-                    "Task guardrail returned None as result. This is not allowed."
-                )
-
-            if isinstance(guardrail_result.result, str):
-                task_output.raw = guardrail_result.result
-                pydantic_output, json_output = self._export_output(
-                    guardrail_result.result
-                )
-                task_output.pydantic = pydantic_output
-                task_output.json_dict = json_output
-            elif isinstance(guardrail_result.result, TaskOutput):
-                task_output = guardrail_result.result
-
-        self.output = task_output
-        self.end_time = datetime.datetime.now()
-
-        if self.callback:
-            self.callback(self.output)
-
-        crew = self.agent.crew  # type: ignore[union-attr]
-        if crew and crew.task_callback and crew.task_callback != self.callback:
-            crew.task_callback(self.output)
-
-        if self._execution_span:
-            self._telemetry.task_ended(self._execution_span, self, agent.crew)
-            self._execution_span = None
-
-        if self.output_file:
-            content = (
-                json_output
-                if json_output
-                else pydantic_output.model_dump_json()
-                if pydantic_output
-                else result
-            )
-            self._save_file(content)
-
-        return task_output
+                self._save_file(content)
+            crewai_event_bus.emit(self, TaskCompletedEvent(output=task_output))
+            return task_output
+        except Exception as e:
+            self.end_time = datetime.datetime.now()
+            crewai_event_bus.emit(self, TaskFailedEvent(error=str(e)))
+            raise e  # Re-raise the exception after emitting the event
 
     def prompt(self) -> str:
         """Prompt the task.
@@ -716,10 +720,9 @@ class Task(BaseModel):
                     file.write(str(result))
         except (OSError, IOError) as e:
             raise RuntimeError(
-                "\n".join([
-                    f"Failed to save output file: {e}",
-                    FILEWRITER_RECOMMENDATION
-                ])
+                "\n".join(
+                    [f"Failed to save output file: {e}", FILEWRITER_RECOMMENDATION]
+                )
             )
         return None
 

src/crewai/tools/tool_usage.py

@@ -1,7 +1,7 @@
 import ast
-import datetime
 import json
 import time
+from datetime import datetime
 from difflib import SequenceMatcher
 from json import JSONDecodeError
 from textwrap import dedent
@@ -10,20 +10,22 @@ from typing import Any, Dict, List, Optional, Union
 import json5
 from json_repair import repair_json
 
-import crewai.utilities.events as events
 from crewai.agents.tools_handler import ToolsHandler
 from crewai.task import Task
 from crewai.telemetry import Telemetry
 from crewai.tools import BaseTool
 from crewai.tools.structured_tool import CrewStructuredTool
 from crewai.tools.tool_calling import InstructorToolCalling, ToolCalling
-from crewai.tools.tool_usage_events import ToolUsageError, ToolUsageFinished
 from crewai.utilities import I18N, Converter, ConverterError, Printer
+from crewai.utilities.datetime_compat import UTC
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
+from crewai.utilities.events.tool_usage_events import (
+    ToolSelectionErrorEvent,
+    ToolUsageErrorEvent,
+    ToolUsageFinishedEvent,
+    ToolValidateInputErrorEvent,
+)
 
-try:
-    import agentops  # type: ignore
-except ImportError:
-    agentops = None
 OPENAI_BIGGER_MODELS = [
     "gpt-4",
     "gpt-4o",
@@ -116,7 +118,10 @@ class ToolUsage:
                 self._printer.print(content=f"\n\n{error}\n", color="red")
             return error
 
-        if isinstance(tool, CrewStructuredTool) and tool.name == self._i18n.tools("add_image")["name"]:  # type: ignore
+        if (
+            isinstance(tool, CrewStructuredTool)
+            and tool.name == self._i18n.tools("add_image")["name"]  # type: ignore
+        ):
             try:
                 result = self._use(tool_string=tool_string, tool=tool, calling=calling)
                 return result
@@ -136,7 +141,6 @@ class ToolUsage:
         tool: Any,
         calling: Union[ToolCalling, InstructorToolCalling],
     ) -> str:  # TODO: Fix this return type
-        tool_event = agentops.ToolEvent(name=calling.tool_name) if agentops else None  # type: ignore
         if self._check_tool_repeated_usage(calling=calling):  # type: ignore # _check_tool_repeated_usage of "ToolUsage" does not return a value (it only ever returns None)
             try:
                 result = self._i18n.errors("task_repeated_usage").format(
@@ -154,6 +158,7 @@ class ToolUsage:
                 self.task.increment_tools_errors()
 
         started_at = time.time()
+        started_at_trace = datetime.now(UTC)
         from_cache = False
 
         result = None  # type: ignore # Incompatible types in assignment (expression has type "None", variable has type "str")
@@ -181,7 +186,9 @@ class ToolUsage:
 
                 if calling.arguments:
                     try:
-                        acceptable_args = tool.args_schema.model_json_schema()["properties"].keys()  # type: ignore
+                        acceptable_args = tool.args_schema.model_json_schema()[
+                            "properties"
+                        ].keys()  # type: ignore
                         arguments = {
                             k: v
                             for k, v in calling.arguments.items()
@@ -202,7 +209,7 @@ class ToolUsage:
                         error=e, tool=tool.name, tool_inputs=tool.description
                     )
                     error = ToolUsageErrorException(
-                        f'\n{error_message}.\nMoving on then. {self._i18n.slice("format").format(tool_names=self.tools_names)}'
+                        f"\n{error_message}.\nMoving on then. {self._i18n.slice('format').format(tool_names=self.tools_names)}"
                     ).message
                     self.task.increment_tools_errors()
                     if self.agent.verbose:
@@ -212,10 +219,6 @@ class ToolUsage:
                     return error  # type: ignore # No return value expected
 
                 self.task.increment_tools_errors()
-                if agentops:
-                    agentops.record(
-                        agentops.ErrorEvent(exception=e, trigger_event=tool_event)
-                    )
                 return self.use(calling=calling, tool_string=tool_string)  # type: ignore # No return value expected
 
             if self.tools_handler:
@@ -231,9 +234,6 @@ class ToolUsage:
                 self.tools_handler.on_tool_use(
                     calling=calling, output=result, should_cache=should_cache
                 )
-
-        if agentops:
-            agentops.record(tool_event)
         self._telemetry.tool_usage(
             llm=self.function_calling_llm,
             tool_name=tool.name,
@@ -244,6 +244,7 @@ class ToolUsage:
             "result": result,
             "tool_name": tool.name,
             "tool_args": calling.arguments,
+            "start_time": started_at_trace,
         }
 
         self.on_tool_use_finished(
@@ -308,14 +309,33 @@ class ToolUsage:
             ):
                 return tool
         self.task.increment_tools_errors()
+        tool_selection_data = {
+            "agent_key": self.agent.key,
+            "agent_role": self.agent.role,
+            "tool_name": tool_name,
+            "tool_args": {},
+            "tool_class": self.tools_description,
+        }
         if tool_name and tool_name != "":
-            raise Exception(
-                f"Action '{tool_name}' don't exist, these are the only available Actions:\n{self.tools_description}"
+            error = f"Action '{tool_name}' don't exist, these are the only available Actions:\n{self.tools_description}"
+            crewai_event_bus.emit(
+                self,
+                ToolSelectionErrorEvent(
+                    **tool_selection_data,
+                    error=error,
+                ),
             )
+            raise Exception(error)
         else:
-            raise Exception(
-                f"I forgot the Action name, these are the only available Actions: {self.tools_description}"
+            error = f"I forgot the Action name, these are the only available Actions: {self.tools_description}"
+            crewai_event_bus.emit(
+                self,
+                ToolSelectionErrorEvent(
+                    **tool_selection_data,
+                    error=error,
+                ),
             )
+            raise Exception(error)
 
     def _render(self) -> str:
         """Render the tool name and description in plain text."""
@@ -368,7 +388,7 @@ class ToolUsage:
                 raise
             else:
                 return ToolUsageErrorException(
-                    f'{self._i18n.errors("tool_arguments_error")}'
+                    f"{self._i18n.errors('tool_arguments_error')}"
                 )
 
         if not isinstance(arguments, dict):
@@ -376,7 +396,7 @@ class ToolUsage:
                 raise
             else:
                 return ToolUsageErrorException(
-                    f'{self._i18n.errors("tool_arguments_error")}'
+                    f"{self._i18n.errors('tool_arguments_error')}"
                 )
 
         return ToolCalling(
@@ -404,7 +424,7 @@ class ToolUsage:
                 if self.agent.verbose:
                     self._printer.print(content=f"\n\n{e}\n", color="red")
                 return ToolUsageErrorException(  # type: ignore # Incompatible return value type (got "ToolUsageErrorException", expected "ToolCalling | InstructorToolCalling")
-                    f'{self._i18n.errors("tool_usage_error").format(error=e)}\nMoving on then. {self._i18n.slice("format").format(tool_names=self.tools_names)}'
+                    f"{self._i18n.errors('tool_usage_error').format(error=e)}\nMoving on then. {self._i18n.slice('format').format(tool_names=self.tools_names)}"
                 )
             return self._tool_calling(tool_string)
 
@@ -451,18 +471,33 @@ class ToolUsage:
             if isinstance(arguments, dict):
                 return arguments
         except Exception as e:
-            self._printer.print(content=f"Failed to repair JSON: {e}", color="red")
+            error = f"Failed to repair JSON: {e}"
+            self._printer.print(content=error, color="red")
 
-        # If all parsing attempts fail, raise an error
-        raise Exception(
+        error_message = (
             "Tool input must be a valid dictionary in JSON or Python literal format"
         )
+        self._emit_validate_input_error(error_message)
+        # If all parsing attempts fail, raise an error
+        raise Exception(error_message)
+
+    def _emit_validate_input_error(self, final_error: str):
+        tool_selection_data = {
+            "agent_key": self.agent.key,
+            "agent_role": self.agent.role,
+            "tool_name": self.action.tool,
+            "tool_args": str(self.action.tool_input),
+            "tool_class": self.__class__.__name__,
+        }
+
+        crewai_event_bus.emit(
+            self,
+            ToolValidateInputErrorEvent(**tool_selection_data, error=final_error),
+        )
 
     def on_tool_error(self, tool: Any, tool_calling: ToolCalling, e: Exception) -> None:
         event_data = self._prepare_event_data(tool, tool_calling)
-        events.emit(
-            source=self, event=ToolUsageError(**{**event_data, "error": str(e)})
-        )
+        crewai_event_bus.emit(self, ToolUsageErrorEvent(**{**event_data, "error": e}))
 
     def on_tool_use_finished(
         self, tool: Any, tool_calling: ToolCalling, from_cache: bool, started_at: float
@@ -471,12 +506,12 @@ class ToolUsage:
         event_data = self._prepare_event_data(tool, tool_calling)
         event_data.update(
             {
-                "started_at": datetime.datetime.fromtimestamp(started_at),
-                "finished_at": datetime.datetime.fromtimestamp(finished_at),
+                "started_at": datetime.fromtimestamp(started_at),
+                "finished_at": datetime.fromtimestamp(finished_at),
                 "from_cache": from_cache,
             }
         )
-        events.emit(source=self, event=ToolUsageFinished(**event_data))
+        crewai_event_bus.emit(self, ToolUsageFinishedEvent(**event_data))
 
     def _prepare_event_data(self, tool: Any, tool_calling: ToolCalling) -> dict:
         return {

src/crewai/tools/tool_usage_events.py

@@ -1,24 +0,0 @@
-from datetime import datetime
-from typing import Any, Dict
-
-from pydantic import BaseModel
-
-
-class ToolUsageEvent(BaseModel):
-    agent_key: str
-    agent_role: str
-    tool_name: str
-    tool_args: Dict[str, Any]
-    tool_class: str
-    run_attempts: int | None = None
-    delegations: int | None = None
-
-
-class ToolUsageFinished(ToolUsageEvent):
-    started_at: datetime
-    finished_at: datetime
-    from_cache: bool = False
-
-
-class ToolUsageError(ToolUsageEvent):
-    error: str

src/crewai/traces/context.py

@@ -0,0 +1,39 @@
+from contextlib import contextmanager
+from contextvars import ContextVar
+from typing import Generator
+
+
+class TraceContext:
+    """Maintains the current trace context throughout the execution stack.
+
+    This class provides a context manager for tracking trace execution across
+    async and sync code paths using ContextVars.
+    """
+
+    _context: ContextVar = ContextVar("trace_context", default=None)
+
+    @classmethod
+    def get_current(cls):
+        """Get the current trace context.
+
+        Returns:
+            Optional[UnifiedTraceController]: The current trace controller or None if not set.
+        """
+        return cls._context.get()
+
+    @classmethod
+    @contextmanager
+    def set_current(cls, trace):
+        """Set the current trace context within a context manager.
+
+        Args:
+            trace: The trace controller to set as current.
+
+        Yields:
+            UnifiedTraceController: The current trace controller.
+        """
+        token = cls._context.set(trace)
+        try:
+            yield trace
+        finally:
+            cls._context.reset(token)

src/crewai/traces/enums.py

@@ -0,0 +1,19 @@
+from enum import Enum
+
+
+class TraceType(Enum):
+    LLM_CALL = "llm_call"
+    TOOL_CALL = "tool_call"
+    FLOW_STEP = "flow_step"
+    START_CALL = "start_call"
+
+
+class RunType(Enum):
+    KICKOFF = "kickoff"
+    TRAIN = "train"
+    TEST = "test"
+
+
+class CrewType(Enum):
+    CREW = "crew"
+    FLOW = "flow"

src/crewai/traces/models.py

@@ -0,0 +1,89 @@
+from datetime import datetime
+from typing import Any, Dict, List, Optional
+
+from pydantic import BaseModel, Field
+
+
+class ToolCall(BaseModel):
+    """Model representing a tool call during execution"""
+
+    name: str
+    arguments: Dict[str, Any]
+    output: str
+    start_time: datetime
+    end_time: Optional[datetime] = None
+    latency_ms: Optional[int] = None
+    error: Optional[str] = None
+
+
+class LLMRequest(BaseModel):
+    """Model representing the LLM request details"""
+
+    model: str
+    messages: List[Dict[str, str]]
+    temperature: Optional[float] = None
+    max_tokens: Optional[int] = None
+    stop_sequences: Optional[List[str]] = None
+    additional_params: Dict[str, Any] = Field(default_factory=dict)
+
+
+class LLMResponse(BaseModel):
+    """Model representing the LLM response details"""
+
+    content: str
+    finish_reason: Optional[str] = None
+
+
+class FlowStepIO(BaseModel):
+    """Model representing flow step input/output details"""
+
+    function_name: str
+    inputs: Dict[str, Any] = Field(default_factory=dict)
+    outputs: Any
+    metadata: Dict[str, Any] = Field(default_factory=dict)
+
+
+class CrewTrace(BaseModel):
+    """Model for tracking detailed information about LLM interactions and Flow steps"""
+
+    deployment_instance_id: Optional[str] = Field(
+        description="ID of the deployment instance"
+    )
+    trace_id: str = Field(description="Unique identifier for this trace")
+    run_id: str = Field(description="Identifier for the execution run")
+    agent_role: Optional[str] = Field(description="Role of the agent")
+    task_id: Optional[str] = Field(description="ID of the current task being executed")
+    task_name: Optional[str] = Field(description="Name of the current task")
+    task_description: Optional[str] = Field(
+        description="Description of the current task"
+    )
+    trace_type: str = Field(description="Type of the trace")
+    crew_type: str = Field(description="Type of the crew")
+    run_type: str = Field(description="Type of the run")
+
+    # Timing information
+    start_time: Optional[datetime] = None
+    end_time: Optional[datetime] = None
+    latency_ms: Optional[int] = None
+
+    # Request/Response for LLM calls
+    request: Optional[LLMRequest] = None
+    response: Optional[LLMResponse] = None
+
+    # Input/Output for Flow steps
+    flow_step: Optional[FlowStepIO] = None
+
+    # Tool usage
+    tool_calls: List[ToolCall] = Field(default_factory=list)
+
+    # Metrics
+    tokens_used: Optional[int] = None
+    prompt_tokens: Optional[int] = None
+    completion_tokens: Optional[int] = None
+    cost: Optional[float] = None
+
+    # Additional metadata
+    status: str = "running"  # running, completed, error
+    error: Optional[str] = None
+    metadata: Dict[str, Any] = Field(default_factory=dict)
+    tags: List[str] = Field(default_factory=list)

src/crewai/traces/unified_trace_controller.py

@@ -0,0 +1,544 @@
+import inspect
+import os
+from datetime import datetime
+from functools import wraps
+from typing import Any, Awaitable, Callable, Dict, List, Optional
+from uuid import uuid4
+
+from crewai.traces.context import TraceContext
+from crewai.traces.enums import CrewType, RunType, TraceType
+from crewai.traces.models import (
+    CrewTrace,
+    FlowStepIO,
+    LLMRequest,
+    LLMResponse,
+    ToolCall,
+)
+from crewai.utilities.datetime_compat import UTC
+
+
+class UnifiedTraceController:
+    """Controls and manages trace execution and recording.
+
+    This class handles the lifecycle of traces including creation, execution tracking,
+    and recording of results for various types of operations (LLM calls, tool calls, flow steps).
+    """
+
+    _task_traces: Dict[str, List["UnifiedTraceController"]] = {}
+
+    def __init__(
+        self,
+        trace_type: TraceType,
+        run_type: RunType,
+        crew_type: CrewType,
+        run_id: str,
+        deployment_instance_id: str = os.environ.get(
+            "CREWAI_DEPLOYMENT_INSTANCE_ID", ""
+        ),
+        parent_trace_id: Optional[str] = None,
+        agent_role: Optional[str] = "unknown",
+        task_name: Optional[str] = None,
+        task_description: Optional[str] = None,
+        task_id: Optional[str] = None,
+        flow_step: Dict[str, Any] = {},
+        tool_calls: List[ToolCall] = [],
+        **context: Any,
+    ) -> None:
+        """Initialize a new trace controller.
+
+        Args:
+            trace_type: Type of trace being recorded.
+            run_type: Type of run being executed.
+            crew_type: Type of crew executing the trace.
+            run_id: Unique identifier for the run.
+            deployment_instance_id: Optional deployment instance identifier.
+            parent_trace_id: Optional parent trace identifier for nested traces.
+            agent_role: Role of the agent executing the trace.
+            task_name: Optional name of the task being executed.
+            task_description: Optional description of the task.
+            task_id: Optional unique identifier for the task.
+            flow_step: Optional flow step information.
+            tool_calls: Optional list of tool calls made during execution.
+            **context: Additional context parameters.
+        """
+        self.trace_id = str(uuid4())
+        self.run_id = run_id
+        self.parent_trace_id = parent_trace_id
+        self.trace_type = trace_type
+        self.run_type = run_type
+        self.crew_type = crew_type
+        self.context = context
+        self.agent_role = agent_role
+        self.task_name = task_name
+        self.task_description = task_description
+        self.task_id = task_id
+        self.deployment_instance_id = deployment_instance_id
+        self.children: List[Dict[str, Any]] = []
+        self.start_time: Optional[datetime] = None
+        self.end_time: Optional[datetime] = None
+        self.error: Optional[str] = None
+        self.tool_calls = tool_calls
+        self.flow_step = flow_step
+        self.status: str = "running"
+
+        # Add trace to task's trace collection if task_id is present
+        if task_id:
+            self._add_to_task_traces()
+
+    def _add_to_task_traces(self) -> None:
+        """Add this trace to the task's trace collection."""
+        if not hasattr(UnifiedTraceController, "_task_traces"):
+            UnifiedTraceController._task_traces = {}
+
+        if self.task_id is None:
+            return
+
+        if self.task_id not in UnifiedTraceController._task_traces:
+            UnifiedTraceController._task_traces[self.task_id] = []
+
+        UnifiedTraceController._task_traces[self.task_id].append(self)
+
+    @classmethod
+    def get_task_traces(cls, task_id: str) -> List["UnifiedTraceController"]:
+        """Get all traces for a specific task.
+
+        Args:
+            task_id: The ID of the task to get traces for
+
+        Returns:
+            List of traces associated with the task
+        """
+        return cls._task_traces.get(task_id, [])
+
+    @classmethod
+    def clear_task_traces(cls, task_id: str) -> None:
+        """Clear traces for a specific task.
+
+        Args:
+            task_id: The ID of the task to clear traces for
+        """
+        if hasattr(cls, "_task_traces") and task_id in cls._task_traces:
+            del cls._task_traces[task_id]
+
+    def _get_current_trace(self) -> "UnifiedTraceController":
+        return TraceContext.get_current()
+
+    def start_trace(self) -> "UnifiedTraceController":
+        """Start the trace execution.
+
+        Returns:
+            UnifiedTraceController: Self for method chaining.
+        """
+        self.start_time = datetime.now(UTC)
+        return self
+
+    def end_trace(self, result: Any = None, error: Optional[str] = None) -> None:
+        """End the trace execution and record results.
+
+        Args:
+            result: Optional result from the trace execution.
+            error: Optional error message if the trace failed.
+        """
+        self.end_time = datetime.now(UTC)
+        self.status = "error" if error else "completed"
+        self.error = error
+        self._record_trace(result)
+
+    def add_child_trace(self, child_trace: Dict[str, Any]) -> None:
+        """Add a child trace to this trace's execution history.
+
+        Args:
+            child_trace: The child trace information to add.
+        """
+        self.children.append(child_trace)
+
+    def to_crew_trace(self) -> CrewTrace:
+        """Convert to CrewTrace format for storage.
+
+        Returns:
+            CrewTrace: The trace data in CrewTrace format.
+        """
+        latency_ms = None
+
+        if self.tool_calls and hasattr(self.tool_calls[0], "start_time"):
+            self.start_time = self.tool_calls[0].start_time
+
+        if self.start_time and self.end_time:
+            latency_ms = int((self.end_time - self.start_time).total_seconds() * 1000)
+
+        request = None
+        response = None
+        flow_step_obj = None
+
+        if self.trace_type in [TraceType.LLM_CALL, TraceType.TOOL_CALL]:
+            request = LLMRequest(
+                model=self.context.get("model", "unknown"),
+                messages=self.context.get("messages", []),
+                temperature=self.context.get("temperature"),
+                max_tokens=self.context.get("max_tokens"),
+                stop_sequences=self.context.get("stop_sequences"),
+            )
+            if "response" in self.context:
+                response = LLMResponse(
+                    content=self.context["response"].get("content", ""),
+                    finish_reason=self.context["response"].get("finish_reason"),
+                )
+
+        elif self.trace_type == TraceType.FLOW_STEP:
+            flow_step_obj = FlowStepIO(
+                function_name=self.flow_step.get("function_name", "unknown"),
+                inputs=self.flow_step.get("inputs", {}),
+                outputs={"result": self.context.get("response")},
+                metadata=self.flow_step.get("metadata", {}),
+            )
+
+        return CrewTrace(
+            deployment_instance_id=self.deployment_instance_id,
+            trace_id=self.trace_id,
+            task_id=self.task_id,
+            run_id=self.run_id,
+            agent_role=self.agent_role,
+            task_name=self.task_name,
+            task_description=self.task_description,
+            trace_type=self.trace_type.value,
+            crew_type=self.crew_type.value,
+            run_type=self.run_type.value,
+            start_time=self.start_time,
+            end_time=self.end_time,
+            latency_ms=latency_ms,
+            request=request,
+            response=response,
+            flow_step=flow_step_obj,
+            tool_calls=self.tool_calls,
+            tokens_used=self.context.get("tokens_used"),
+            prompt_tokens=self.context.get("prompt_tokens"),
+            completion_tokens=self.context.get("completion_tokens"),
+            status=self.status,
+            error=self.error,
+        )
+
+    def _record_trace(self, result: Any = None) -> None:
+        """Record the trace.
+
+        This method is called when a trace is completed. It ensures the trace
+        is properly recorded and associated with its task if applicable.
+
+        Args:
+            result: Optional result to include in the trace
+        """
+        if result:
+            self.context["response"] = result
+
+        # Add to task traces if this trace belongs to a task
+        if self.task_id:
+            self._add_to_task_traces()
+
+
+def should_trace() -> bool:
+    """Check if tracing is enabled via environment variable."""
+    return os.getenv("CREWAI_ENABLE_TRACING", "false").lower() == "true"
+
+
+# Crew main trace
+def init_crew_main_trace(func: Callable[..., Any]) -> Callable[..., Any]:
+    """Decorator to initialize and track the main crew execution trace.
+
+    This decorator sets up the trace context for the main crew execution,
+    handling both synchronous and asynchronous crew operations.
+
+    Args:
+        func: The crew function to be traced.
+
+    Returns:
+        Wrapped function that creates and manages the main crew trace context.
+    """
+
+    @wraps(func)
+    def wrapper(self: Any, *args: Any, **kwargs: Any) -> Any:
+        if not should_trace():
+            return func(self, *args, **kwargs)
+
+        trace = build_crew_main_trace(self)
+        with TraceContext.set_current(trace):
+            try:
+                return func(self, *args, **kwargs)
+            except Exception as e:
+                trace.end_trace(error=str(e))
+                raise
+
+    return wrapper
+
+
+def build_crew_main_trace(self: Any) -> "UnifiedTraceController":
+    """Build the main trace controller for a crew execution.
+
+    This function creates a trace controller configured for the main crew execution,
+    handling different run types (kickoff, test, train) and maintaining context.
+
+    Args:
+        self: The crew instance.
+
+    Returns:
+        UnifiedTraceController: The configured trace controller for the crew.
+    """
+    run_type = RunType.KICKOFF
+    if hasattr(self, "_test") and self._test:
+        run_type = RunType.TEST
+    elif hasattr(self, "_train") and self._train:
+        run_type = RunType.TRAIN
+
+    current_trace = TraceContext.get_current()
+
+    trace = UnifiedTraceController(
+        trace_type=TraceType.LLM_CALL,
+        run_type=run_type,
+        crew_type=current_trace.crew_type if current_trace else CrewType.CREW,
+        run_id=current_trace.run_id if current_trace else str(self.id),
+        parent_trace_id=current_trace.trace_id if current_trace else None,
+    )
+    return trace
+
+
+# Flow main trace
+def init_flow_main_trace(
+    func: Callable[..., Awaitable[Any]],
+) -> Callable[..., Awaitable[Any]]:
+    """Decorator to initialize and track the main flow execution trace.
+
+    Args:
+        func: The async flow function to be traced.
+
+    Returns:
+        Wrapped async function that creates and manages the main flow trace context.
+    """
+
+    @wraps(func)
+    async def wrapper(self: Any, *args: Any, **kwargs: Any) -> Any:
+        if not should_trace():
+            return await func(self, *args, **kwargs)
+
+        trace = build_flow_main_trace(self, *args, **kwargs)
+        with TraceContext.set_current(trace):
+            try:
+                return await func(self, *args, **kwargs)
+            except Exception:
+                raise
+
+    return wrapper
+
+
+def build_flow_main_trace(
+    self: Any, *args: Any, **kwargs: Any
+) -> "UnifiedTraceController":
+    """Build the main trace controller for a flow execution.
+
+    Args:
+        self: The flow instance.
+        *args: Variable positional arguments.
+        **kwargs: Variable keyword arguments.
+
+    Returns:
+        UnifiedTraceController: The configured trace controller for the flow.
+    """
+    current_trace = TraceContext.get_current()
+    trace = UnifiedTraceController(
+        trace_type=TraceType.FLOW_STEP,
+        run_id=current_trace.run_id if current_trace else str(self.flow_id),
+        parent_trace_id=current_trace.trace_id if current_trace else None,
+        crew_type=CrewType.FLOW,
+        run_type=RunType.KICKOFF,
+        context={
+            "crew_name": self.__class__.__name__,
+            "inputs": kwargs.get("inputs", {}),
+            "agents": [],
+            "tasks": [],
+        },
+    )
+    return trace
+
+
+# Flow step trace
+def trace_flow_step(
+    func: Callable[..., Awaitable[Any]],
+) -> Callable[..., Awaitable[Any]]:
+    """Decorator to trace individual flow step executions.
+
+    Args:
+        func: The async flow step function to be traced.
+
+    Returns:
+        Wrapped async function that creates and manages the flow step trace context.
+    """
+
+    @wraps(func)
+    async def wrapper(
+        self: Any,
+        method_name: str,
+        method: Callable[..., Any],
+        *args: Any,
+        **kwargs: Any,
+    ) -> Any:
+        if not should_trace():
+            return await func(self, method_name, method, *args, **kwargs)
+
+        trace = build_flow_step_trace(self, method_name, method, *args, **kwargs)
+        with TraceContext.set_current(trace):
+            trace.start_trace()
+            try:
+                result = await func(self, method_name, method, *args, **kwargs)
+                trace.end_trace(result=result)
+                return result
+            except Exception as e:
+                trace.end_trace(error=str(e))
+                raise
+
+    return wrapper
+
+
+def build_flow_step_trace(
+    self: Any, method_name: str, method: Callable[..., Any], *args: Any, **kwargs: Any
+) -> "UnifiedTraceController":
+    """Build a trace controller for an individual flow step.
+
+    Args:
+        self: The flow instance.
+        method_name: Name of the method being executed.
+        method: The actual method being executed.
+        *args: Variable positional arguments.
+        **kwargs: Variable keyword arguments.
+
+    Returns:
+        UnifiedTraceController: The configured trace controller for the flow step.
+    """
+    current_trace = TraceContext.get_current()
+
+    # Get method signature
+    sig = inspect.signature(method)
+    params = list(sig.parameters.values())
+
+    # Create inputs dictionary mapping parameter names to values
+    method_params = [p for p in params if p.name != "self"]
+    inputs: Dict[str, Any] = {}
+
+    # Map positional args to their parameter names
+    for i, param in enumerate(method_params):
+        if i < len(args):
+            inputs[param.name] = args[i]
+
+    # Add keyword arguments
+    inputs.update(kwargs)
+
+    trace = UnifiedTraceController(
+        trace_type=TraceType.FLOW_STEP,
+        run_type=current_trace.run_type if current_trace else RunType.KICKOFF,
+        crew_type=current_trace.crew_type if current_trace else CrewType.FLOW,
+        run_id=current_trace.run_id if current_trace else str(self.flow_id),
+        parent_trace_id=current_trace.trace_id if current_trace else None,
+        flow_step={
+            "function_name": method_name,
+            "inputs": inputs,
+            "metadata": {
+                "crew_name": self.__class__.__name__,
+            },
+        },
+    )
+    return trace
+
+
+# LLM trace
+def trace_llm_call(func: Callable[..., Any]) -> Callable[..., Any]:
+    """Decorator to trace LLM calls.
+
+    Args:
+        func: The function to trace.
+
+    Returns:
+        Wrapped function that creates and manages the LLM call trace context.
+    """
+
+    @wraps(func)
+    def wrapper(self: Any, *args: Any, **kwargs: Any) -> Any:
+        if not should_trace():
+            return func(self, *args, **kwargs)
+
+        trace = build_llm_trace(self, *args, **kwargs)
+        with TraceContext.set_current(trace):
+            trace.start_trace()
+            try:
+                response = func(self, *args, **kwargs)
+                # Extract relevant data from response
+                trace_response = {
+                    "content": response["choices"][0]["message"]["content"],
+                    "finish_reason": response["choices"][0].get("finish_reason"),
+                }
+
+                # Add usage metrics to context
+                if "usage" in response:
+                    trace.context["tokens_used"] = response["usage"].get(
+                        "total_tokens", 0
+                    )
+                    trace.context["prompt_tokens"] = response["usage"].get(
+                        "prompt_tokens", 0
+                    )
+                    trace.context["completion_tokens"] = response["usage"].get(
+                        "completion_tokens", 0
+                    )
+
+                trace.end_trace(trace_response)
+                return response
+            except Exception as e:
+                trace.end_trace(error=str(e))
+                raise
+
+    return wrapper
+
+
+def build_llm_trace(
+    self: Any, params: Dict[str, Any], *args: Any, **kwargs: Any
+) -> Any:
+    """Build a trace controller for an LLM call.
+
+    Args:
+        self: The LLM instance.
+        params: The parameters for the LLM call.
+        *args: Variable positional arguments.
+        **kwargs: Variable keyword arguments.
+
+    Returns:
+        UnifiedTraceController: The configured trace controller for the LLM call.
+    """
+    current_trace = TraceContext.get_current()
+    agent, task = self._get_execution_context()
+
+    # Get new messages and tool results
+    new_messages = self._get_new_messages(params.get("messages", []))
+    new_tool_results = self._get_new_tool_results(agent)
+
+    # Create trace context
+    trace = UnifiedTraceController(
+        trace_type=TraceType.TOOL_CALL if new_tool_results else TraceType.LLM_CALL,
+        crew_type=current_trace.crew_type if current_trace else CrewType.CREW,
+        run_type=current_trace.run_type if current_trace else RunType.KICKOFF,
+        run_id=current_trace.run_id if current_trace else str(uuid4()),
+        parent_trace_id=current_trace.trace_id if current_trace else None,
+        agent_role=agent.role if agent else "unknown",
+        task_id=str(task.id) if task else None,
+        task_name=task.name if task else None,
+        task_description=task.description if task else None,
+        model=self.model,
+        messages=new_messages,
+        temperature=self.temperature,
+        max_tokens=self.max_tokens,
+        stop_sequences=self.stop,
+        tool_calls=[
+            ToolCall(
+                name=result["tool_name"],
+                arguments=result["tool_args"],
+                output=str(result["result"]),
+                start_time=result.get("start_time", ""),
+                end_time=datetime.now(UTC),
+            )
+            for result in new_tool_results
+        ],
+    )
+    return trace

src/crewai/utilities/__init__.py

@@ -1,4 +1,5 @@
 from .converter import Converter, ConverterError
+from .datetime_compat import UTC
 from .file_handler import FileHandler
 from .i18n import I18N
 from .internal_instructor import InternalInstructor
@@ -22,6 +23,7 @@ __all__ = [
     "Printer",
     "Prompts",
     "RPMController",
+    "UTC",
     "YamlParser",
     "LLMContextLengthExceededException",
     "EmbeddingConfigurator",

src/crewai/utilities/constants.py

@@ -4,3 +4,4 @@ DEFAULT_SCORE_THRESHOLD = 0.35
 KNOWLEDGE_DIRECTORY = "knowledge"
 MAX_LLM_RETRY = 3
 MAX_FILE_NAME_LENGTH = 255
+EMITTER_COLOR = "bold_blue"

src/crewai/utilities/datetime_compat.py

@@ -0,0 +1,23 @@
+"""Compatibility module for datetime functionality across Python versions.
+
+This module provides timezone constants that work consistently across different
+Python versions, particularly focusing on maintaining compatibility between
+Python 3.10 and newer versions.
+
+Notes:
+    - In Python 3.10, datetime.UTC is not available, so we use timezone.utc
+    - In Python 3.11+, this provides equivalent functionality to datetime.UTC
+    - This implementation maintains consistent behavior across versions for
+      timezone-aware datetime operations
+    - No known limitations or edge cases between versions
+    - Safe to use with DST transitions and leap years
+    - Maintains exact timezone offset (always UTC+00:00)
+
+Example:
+    >>> from datetime import datetime
+    >>> from crewai.utilities.datetime_compat import UTC
+    >>> dt = datetime.now(UTC)  # Creates timezone-aware datetime with UTC
+"""
+from datetime import timezone
+
+UTC = timezone.utc  # Equivalent to datetime.UTC (Python 3.11+)

src/crewai/utilities/evaluators/crew_evaluator_handler.py

@@ -1,11 +1,12 @@
 from collections import defaultdict
 
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, InstanceOf
 from rich.box import HEAVY_EDGE
 from rich.console import Console
 from rich.table import Table
 
 from crewai.agent import Agent
+from crewai.llm import LLM
 from crewai.task import Task
 from crewai.tasks.task_output import TaskOutput
 from crewai.telemetry import Telemetry
@@ -23,7 +24,7 @@ class CrewEvaluator:
 
     Attributes:
         crew (Crew): The crew of agents to evaluate.
-        openai_model_name (str): The model to use for evaluating the performance of the agents (for now ONLY OpenAI accepted).
+        eval_llm (LLM): Language model instance to use for evaluations
         tasks_scores (defaultdict): A dictionary to store the scores of the agents for each task.
         iteration (int): The current iteration of the evaluation.
     """
@@ -32,9 +33,9 @@ class CrewEvaluator:
     run_execution_times: defaultdict = defaultdict(list)
     iteration: int = 0
 
-    def __init__(self, crew, openai_model_name: str):
+    def __init__(self, crew, eval_llm: InstanceOf[LLM]):
         self.crew = crew
-        self.openai_model_name = openai_model_name
+        self.llm = eval_llm
         self._telemetry = Telemetry()
         self._setup_for_evaluating()
 
@@ -51,7 +52,7 @@ class CrewEvaluator:
             ),
             backstory="Evaluator agent for crew evaluation with precise capabilities to evaluate the performance of the agents in the crew based on the tasks they have performed",
             verbose=False,
-            llm=self.openai_model_name,
+            llm=self.llm,
         )
 
     def _evaluation_task(
@@ -181,7 +182,7 @@ class CrewEvaluator:
                 self.crew,
                 evaluation_result.pydantic.quality,
                 current_task.execution_duration,
-                self.openai_model_name,
+                self.llm.model,
             )
             self.tasks_scores[self.iteration].append(evaluation_result.pydantic.quality)
             self.run_execution_times[self.iteration].append(

src/crewai/utilities/evaluators/task_evaluator.py

@@ -3,19 +3,9 @@ from typing import List
 from pydantic import BaseModel, Field
 
 from crewai.utilities import Converter
+from crewai.utilities.events import TaskEvaluationEvent, crewai_event_bus
 from crewai.utilities.pydantic_schema_parser import PydanticSchemaParser
 
-agentops = None
-try:
-    from agentops import track_agent  # type: ignore
-except ImportError:
-
-    def track_agent(name):
-        def noop(f):
-            return f
-
-        return noop
-
 
 class Entity(BaseModel):
     name: str = Field(description="The name of the entity.")
@@ -48,12 +38,15 @@ class TrainingTaskEvaluation(BaseModel):
     )
 
 
-@track_agent(name="Task Evaluator")
 class TaskEvaluator:
     def __init__(self, original_agent):
         self.llm = original_agent.llm
+        self.original_agent = original_agent
 
     def evaluate(self, task, output) -> TaskEvaluation:
+        crewai_event_bus.emit(
+            self, TaskEvaluationEvent(evaluation_type="task_evaluation")
+        )
         evaluation_query = (
             f"Assess the quality of the task completed based on the description, expected output, and actual results.\n\n"
             f"Task Description:\n{task.description}\n\n"
@@ -90,6 +83,9 @@ class TaskEvaluator:
             - training_data (dict): The training data to be evaluated.
             - agent_id (str): The ID of the agent.
         """
+        crewai_event_bus.emit(
+            self, TaskEvaluationEvent(evaluation_type="training_data_evaluation")
+        )
 
         output_training_data = training_data[agent_id]
         final_aggregated_data = ""

src/crewai/utilities/events.py

@@ -1,44 +0,0 @@
-from functools import wraps
-from typing import Any, Callable, Dict, Generic, List, Type, TypeVar
-
-from pydantic import BaseModel
-
-T = TypeVar("T")
-EVT = TypeVar("EVT", bound=BaseModel)
-
-
-class Emitter(Generic[T, EVT]):
-    _listeners: Dict[Type[EVT], List[Callable]] = {}
-
-    def on(self, event_type: Type[EVT]):
-        def decorator(func: Callable):
-            @wraps(func)
-            def wrapper(*args, **kwargs):
-                return func(*args, **kwargs)
-
-            self._listeners.setdefault(event_type, []).append(wrapper)
-            return wrapper
-
-        return decorator
-
-    def emit(self, source: T, event: EVT) -> None:
-        event_type = type(event)
-        for func in self._listeners.get(event_type, []):
-            func(source, event)
-
-
-default_emitter = Emitter[Any, BaseModel]()
-
-
-def emit(source: Any, event: BaseModel, raise_on_error: bool = False) -> None:
-    try:
-        default_emitter.emit(source, event)
-    except Exception as e:
-        if raise_on_error:
-            raise e
-        else:
-            print(f"Error emitting event: {e}")
-
-
-def on(event_type: Type[BaseModel]) -> Callable:
-    return default_emitter.on(event_type)

src/crewai/utilities/events/__init__.py

@@ -0,0 +1,40 @@
+from .crew_events import (
+    CrewKickoffStartedEvent,
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewTrainStartedEvent,
+    CrewTrainCompletedEvent,
+    CrewTrainFailedEvent,
+    CrewTestStartedEvent,
+    CrewTestCompletedEvent,
+    CrewTestFailedEvent,
+)
+from .agent_events import (
+    AgentExecutionStartedEvent,
+    AgentExecutionCompletedEvent,
+    AgentExecutionErrorEvent,
+)
+from .task_events import TaskStartedEvent, TaskCompletedEvent, TaskFailedEvent, TaskEvaluationEvent
+from .flow_events import (
+    FlowCreatedEvent,
+    FlowStartedEvent,
+    FlowFinishedEvent,
+    FlowPlotEvent,
+    MethodExecutionStartedEvent,
+    MethodExecutionFinishedEvent,
+    MethodExecutionFailedEvent,
+)
+from .crewai_event_bus import CrewAIEventsBus, crewai_event_bus
+from .tool_usage_events import (
+    ToolUsageFinishedEvent,
+    ToolUsageErrorEvent,
+    ToolUsageStartedEvent,
+    ToolExecutionErrorEvent,
+    ToolSelectionErrorEvent,
+    ToolUsageEvent,
+    ToolValidateInputErrorEvent,
+)
+
+# events
+from .event_listener import EventListener
+from .third_party.agentops_listener import agentops_listener

src/crewai/utilities/events/agent_events.py

@@ -0,0 +1,40 @@
+from typing import TYPE_CHECKING, Any, Dict, Optional, Sequence, Union
+
+from crewai.agents.agent_builder.base_agent import BaseAgent
+from crewai.tools.base_tool import BaseTool
+from crewai.tools.structured_tool import CrewStructuredTool
+
+from .base_events import CrewEvent
+
+if TYPE_CHECKING:
+    from crewai.agents.agent_builder.base_agent import BaseAgent
+
+
+class AgentExecutionStartedEvent(CrewEvent):
+    """Event emitted when an agent starts executing a task"""
+
+    agent: BaseAgent
+    task: Any
+    tools: Optional[Sequence[Union[BaseTool, CrewStructuredTool]]]
+    task_prompt: str
+    type: str = "agent_execution_started"
+
+    model_config = {"arbitrary_types_allowed": True}
+
+
+class AgentExecutionCompletedEvent(CrewEvent):
+    """Event emitted when an agent completes executing a task"""
+
+    agent: BaseAgent
+    task: Any
+    output: str
+    type: str = "agent_execution_completed"
+
+
+class AgentExecutionErrorEvent(CrewEvent):
+    """Event emitted when an agent encounters an error during execution"""
+
+    agent: BaseAgent
+    task: Any
+    error: str
+    type: str = "agent_execution_error"

src/crewai/utilities/events/base_event_listener.py

@@ -0,0 +1,14 @@
+from abc import ABC, abstractmethod
+from logging import Logger
+
+from crewai.utilities.events.crewai_event_bus import CrewAIEventsBus, crewai_event_bus
+
+
+class BaseEventListener(ABC):
+    def __init__(self):
+        super().__init__()
+        self.setup_listeners(crewai_event_bus)
+
+    @abstractmethod
+    def setup_listeners(self, crewai_event_bus: CrewAIEventsBus):
+        pass

src/crewai/utilities/events/base_events.py

@@ -0,0 +1,10 @@
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+class CrewEvent(BaseModel):
+    """Base class for all crew events"""
+
+    timestamp: datetime = Field(default_factory=datetime.now)
+    type: str

src/crewai/utilities/events/crew_events.py

@@ -0,0 +1,81 @@
+from typing import Any, Dict, Optional, Union
+
+from pydantic import InstanceOf
+
+from crewai.utilities.events.base_events import CrewEvent
+
+
+class CrewKickoffStartedEvent(CrewEvent):
+    """Event emitted when a crew starts execution"""
+
+    crew_name: Optional[str]
+    inputs: Optional[Dict[str, Any]]
+    type: str = "crew_kickoff_started"
+
+
+class CrewKickoffCompletedEvent(CrewEvent):
+    """Event emitted when a crew completes execution"""
+
+    crew_name: Optional[str]
+    output: Any
+    type: str = "crew_kickoff_completed"
+
+
+class CrewKickoffFailedEvent(CrewEvent):
+    """Event emitted when a crew fails to complete execution"""
+
+    error: str
+    crew_name: Optional[str]
+    type: str = "crew_kickoff_failed"
+
+
+class CrewTrainStartedEvent(CrewEvent):
+    """Event emitted when a crew starts training"""
+
+    crew_name: Optional[str]
+    n_iterations: int
+    filename: str
+    inputs: Optional[Dict[str, Any]]
+    type: str = "crew_train_started"
+
+
+class CrewTrainCompletedEvent(CrewEvent):
+    """Event emitted when a crew completes training"""
+
+    crew_name: Optional[str]
+    n_iterations: int
+    filename: str
+    type: str = "crew_train_completed"
+
+
+class CrewTrainFailedEvent(CrewEvent):
+    """Event emitted when a crew fails to complete training"""
+
+    error: str
+    crew_name: Optional[str]
+    type: str = "crew_train_failed"
+
+
+class CrewTestStartedEvent(CrewEvent):
+    """Event emitted when a crew starts testing"""
+
+    crew_name: Optional[str]
+    n_iterations: int
+    eval_llm: Optional[Union[str, Any]]
+    inputs: Optional[Dict[str, Any]]
+    type: str = "crew_test_started"
+
+
+class CrewTestCompletedEvent(CrewEvent):
+    """Event emitted when a crew completes testing"""
+
+    crew_name: Optional[str]
+    type: str = "crew_test_completed"
+
+
+class CrewTestFailedEvent(CrewEvent):
+    """Event emitted when a crew fails to complete testing"""
+
+    error: str
+    crew_name: Optional[str]
+    type: str = "crew_test_failed"

src/crewai/utilities/events/crewai_event_bus.py

@@ -0,0 +1,113 @@
+import threading
+from contextlib import contextmanager
+from typing import Any, Callable, Dict, List, Type, TypeVar, cast
+
+from blinker import Signal
+
+from crewai.utilities.events.base_events import CrewEvent
+from crewai.utilities.events.event_types import EventTypes
+
+EventT = TypeVar("EventT", bound=CrewEvent)
+
+
+class CrewAIEventsBus:
+    """
+    A singleton event bus that uses blinker signals for event handling.
+    Allows both internal (Flow/Crew) and external event handling.
+    """
+
+    _instance = None
+    _lock = threading.Lock()
+
+    def __new__(cls):
+        if cls._instance is None:
+            with cls._lock:
+                if cls._instance is None:  # prevent race condition
+                    cls._instance = super(CrewAIEventsBus, cls).__new__(cls)
+                    cls._instance._initialize()
+        return cls._instance
+
+    def _initialize(self) -> None:
+        """Initialize the event bus internal state"""
+        self._signal = Signal("crewai_event_bus")
+        self._handlers: Dict[Type[CrewEvent], List[Callable]] = {}
+
+    def on(
+        self, event_type: Type[EventT]
+    ) -> Callable[[Callable[[Any, EventT], None]], Callable[[Any, EventT], None]]:
+        """
+        Decorator to register an event handler for a specific event type.
+
+        Usage:
+            @crewai_event_bus.on(AgentExecutionCompletedEvent)
+            def on_agent_execution_completed(
+                source: Any, event: AgentExecutionCompletedEvent
+            ):
+                print(f"👍 Agent '{event.agent}' completed task")
+                print(f"   Output: {event.output}")
+        """
+
+        def decorator(
+            handler: Callable[[Any, EventT], None],
+        ) -> Callable[[Any, EventT], None]:
+            if event_type not in self._handlers:
+                self._handlers[event_type] = []
+            self._handlers[event_type].append(
+                cast(Callable[[Any, EventT], None], handler)
+            )
+            return handler
+
+        return decorator
+
+    def emit(self, source: Any, event: CrewEvent) -> None:
+        """
+        Emit an event to all registered handlers
+
+        Args:
+            source: The object emitting the event
+            event: The event instance to emit
+        """
+        event_type = type(event)
+        if event_type in self._handlers:
+            for handler in self._handlers[event_type]:
+                handler(source, event)
+        self._signal.send(source, event=event)
+
+    def clear_handlers(self) -> None:
+        """Clear all registered event handlers - useful for testing"""
+        self._handlers.clear()
+
+    def register_handler(
+        self, event_type: Type[EventTypes], handler: Callable[[Any, EventTypes], None]
+    ) -> None:
+        """Register an event handler for a specific event type"""
+        if event_type not in self._handlers:
+            self._handlers[event_type] = []
+        self._handlers[event_type].append(
+            cast(Callable[[Any, EventTypes], None], handler)
+        )
+
+    @contextmanager
+    def scoped_handlers(self):
+        """
+        Context manager for temporary event handling scope.
+        Useful for testing or temporary event handling.
+
+        Usage:
+            with crewai_event_bus.scoped_handlers():
+                @crewai_event_bus.on(CrewKickoffStarted)
+                def temp_handler(source, event):
+                    print("Temporary handler")
+                # Do stuff...
+            # Handlers are cleared after the context
+        """
+        previous_handlers = self._handlers.copy()
+        self._handlers.clear()
+        try:
+            yield
+        finally:
+            self._handlers = previous_handlers
+
+
+# Global instance
+crewai_event_bus = CrewAIEventsBus()

src/crewai/utilities/events/event_listener.py

@@ -0,0 +1,257 @@
+from pydantic import PrivateAttr
+
+from crewai.telemetry.telemetry import Telemetry
+from crewai.utilities import Logger
+from crewai.utilities.constants import EMITTER_COLOR
+from crewai.utilities.events.base_event_listener import BaseEventListener
+
+from .agent_events import AgentExecutionCompletedEvent, AgentExecutionStartedEvent
+from .crew_events import (
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewKickoffStartedEvent,
+    CrewTestCompletedEvent,
+    CrewTestFailedEvent,
+    CrewTestStartedEvent,
+    CrewTrainCompletedEvent,
+    CrewTrainFailedEvent,
+    CrewTrainStartedEvent,
+)
+from .flow_events import (
+    FlowCreatedEvent,
+    FlowFinishedEvent,
+    FlowStartedEvent,
+    MethodExecutionFailedEvent,
+    MethodExecutionFinishedEvent,
+    MethodExecutionStartedEvent,
+)
+from .task_events import TaskCompletedEvent, TaskFailedEvent, TaskStartedEvent
+from .tool_usage_events import (
+    ToolUsageErrorEvent,
+    ToolUsageFinishedEvent,
+    ToolUsageStartedEvent,
+)
+
+
+class EventListener(BaseEventListener):
+    _instance = None
+    _telemetry: Telemetry = PrivateAttr(default_factory=lambda: Telemetry())
+    logger = Logger(verbose=True, default_color=EMITTER_COLOR)
+
+    def __new__(cls):
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+            cls._instance._initialized = False
+        return cls._instance
+
+    def __init__(self):
+        if not hasattr(self, "_initialized") or not self._initialized:
+            super().__init__()
+            self._telemetry = Telemetry()
+            self._telemetry.set_tracer()
+            self._initialized = True
+
+    # ----------- CREW EVENTS -----------
+
+    def setup_listeners(self, crewai_event_bus):
+        @crewai_event_bus.on(CrewKickoffStartedEvent)
+        def on_crew_started(source, event: CrewKickoffStartedEvent):
+            self.logger.log(
+                f"🚀 Crew '{event.crew_name}' started",
+                event.timestamp,
+            )
+            self._telemetry.crew_execution_span(source, event.inputs)
+
+        @crewai_event_bus.on(CrewKickoffCompletedEvent)
+        def on_crew_completed(source, event: CrewKickoffCompletedEvent):
+            final_string_output = event.output.raw
+            self._telemetry.end_crew(source, final_string_output)
+            self.logger.log(
+                f"✅ Crew '{event.crew_name}' completed",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewKickoffFailedEvent)
+        def on_crew_failed(source, event: CrewKickoffFailedEvent):
+            self.logger.log(
+                f"❌ Crew '{event.crew_name}' failed",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTestStartedEvent)
+        def on_crew_test_started(source, event: CrewTestStartedEvent):
+            cloned_crew = source.copy()
+            cloned_crew._telemetry.test_execution_span(
+                cloned_crew,
+                event.n_iterations,
+                event.inputs,
+                event.eval_llm,
+            )
+            self.logger.log(
+                f"🚀 Crew '{event.crew_name}' started test",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTestCompletedEvent)
+        def on_crew_test_completed(source, event: CrewTestCompletedEvent):
+            self.logger.log(
+                f"✅ Crew '{event.crew_name}' completed test",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTestFailedEvent)
+        def on_crew_test_failed(source, event: CrewTestFailedEvent):
+            self.logger.log(
+                f"❌ Crew '{event.crew_name}' failed test",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTrainStartedEvent)
+        def on_crew_train_started(source, event: CrewTrainStartedEvent):
+            self.logger.log(
+                f"📋 Crew '{event.crew_name}' started train",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTrainCompletedEvent)
+        def on_crew_train_completed(source, event: CrewTrainCompletedEvent):
+            self.logger.log(
+                f"✅ Crew '{event.crew_name}' completed train",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(CrewTrainFailedEvent)
+        def on_crew_train_failed(source, event: CrewTrainFailedEvent):
+            self.logger.log(
+                f"❌ Crew '{event.crew_name}' failed train",
+                event.timestamp,
+            )
+
+        # ----------- TASK EVENTS -----------
+
+        @crewai_event_bus.on(TaskStartedEvent)
+        def on_task_started(source, event: TaskStartedEvent):
+            source._execution_span = self._telemetry.task_started(
+                crew=source.agent.crew, task=source
+            )
+            self.logger.log(
+                f"📋 Task started: {source.description}",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(TaskCompletedEvent)
+        def on_task_completed(source, event: TaskCompletedEvent):
+            if source._execution_span:
+                self._telemetry.task_ended(
+                    source._execution_span, source, source.agent.crew
+                )
+            self.logger.log(
+                f"✅ Task completed: {source.description}",
+                event.timestamp,
+            )
+            source._execution_span = None
+
+        @crewai_event_bus.on(TaskFailedEvent)
+        def on_task_failed(source, event: TaskFailedEvent):
+            if source._execution_span:
+                if source.agent and source.agent.crew:
+                    self._telemetry.task_ended(
+                        source._execution_span, source, source.agent.crew
+                    )
+                source._execution_span = None
+            self.logger.log(
+                f"❌ Task failed: {source.description}",
+                event.timestamp,
+            )
+
+        # ----------- AGENT EVENTS -----------
+
+        @crewai_event_bus.on(AgentExecutionStartedEvent)
+        def on_agent_execution_started(source, event: AgentExecutionStartedEvent):
+            self.logger.log(
+                f"🤖 Agent '{event.agent.role}' started task",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(AgentExecutionCompletedEvent)
+        def on_agent_execution_completed(source, event: AgentExecutionCompletedEvent):
+            self.logger.log(
+                f"✅ Agent '{event.agent.role}' completed task",
+                event.timestamp,
+            )
+
+        # ----------- FLOW EVENTS -----------
+
+        @crewai_event_bus.on(FlowCreatedEvent)
+        def on_flow_created(source, event: FlowCreatedEvent):
+            self._telemetry.flow_creation_span(self.__class__.__name__)
+            self.logger.log(
+                f"🌊 Flow Created: '{event.flow_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(FlowStartedEvent)
+        def on_flow_started(source, event: FlowStartedEvent):
+            self._telemetry.flow_execution_span(
+                source.__class__.__name__, list(source._methods.keys())
+            )
+            self.logger.log(
+                f"🤖 Flow Started: '{event.flow_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(FlowFinishedEvent)
+        def on_flow_finished(source, event: FlowFinishedEvent):
+            self.logger.log(
+                f"👍 Flow Finished: '{event.flow_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(MethodExecutionStartedEvent)
+        def on_method_execution_started(source, event: MethodExecutionStartedEvent):
+            self.logger.log(
+                f"🤖 Flow Method Started: '{event.method_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(MethodExecutionFailedEvent)
+        def on_method_execution_failed(source, event: MethodExecutionFailedEvent):
+            self.logger.log(
+                f"❌ Flow Method Failed: '{event.method_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(MethodExecutionFinishedEvent)
+        def on_method_execution_finished(source, event: MethodExecutionFinishedEvent):
+            self.logger.log(
+                f"👍 Flow Method Finished: '{event.method_name}'",
+                event.timestamp,
+            )
+
+        # ----------- TOOL USAGE EVENTS -----------
+
+        @crewai_event_bus.on(ToolUsageStartedEvent)
+        def on_tool_usage_started(source, event: ToolUsageStartedEvent):
+            self.logger.log(
+                f"🤖 Tool Usage Started: '{event.tool_name}'",
+                event.timestamp,
+            )
+
+        @crewai_event_bus.on(ToolUsageFinishedEvent)
+        def on_tool_usage_finished(source, event: ToolUsageFinishedEvent):
+            self.logger.log(
+                f"✅ Tool Usage Finished: '{event.tool_name}'",
+                event.timestamp,
+                #
+            )
+
+        @crewai_event_bus.on(ToolUsageErrorEvent)
+        def on_tool_usage_error(source, event: ToolUsageErrorEvent):
+            self.logger.log(
+                f"❌ Tool Usage Error: '{event.tool_name}'",
+                event.timestamp,
+                #
+            )
+
+
+event_listener = EventListener()

src/crewai/utilities/events/event_types.py

@@ -0,0 +1,61 @@
+from typing import Union
+
+from .agent_events import (
+    AgentExecutionCompletedEvent,
+    AgentExecutionErrorEvent,
+    AgentExecutionStartedEvent,
+)
+from .crew_events import (
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewKickoffStartedEvent,
+    CrewTestCompletedEvent,
+    CrewTestFailedEvent,
+    CrewTestStartedEvent,
+    CrewTrainCompletedEvent,
+    CrewTrainFailedEvent,
+    CrewTrainStartedEvent,
+)
+from .flow_events import (
+    FlowFinishedEvent,
+    FlowStartedEvent,
+    MethodExecutionFailedEvent,
+    MethodExecutionFinishedEvent,
+    MethodExecutionStartedEvent,
+)
+from .task_events import (
+    TaskCompletedEvent,
+    TaskFailedEvent,
+    TaskStartedEvent,
+)
+from .tool_usage_events import (
+    ToolUsageErrorEvent,
+    ToolUsageFinishedEvent,
+    ToolUsageStartedEvent,
+)
+
+EventTypes = Union[
+    CrewKickoffStartedEvent,
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewTestStartedEvent,
+    CrewTestCompletedEvent,
+    CrewTestFailedEvent,
+    CrewTrainStartedEvent,
+    CrewTrainCompletedEvent,
+    CrewTrainFailedEvent,
+    AgentExecutionStartedEvent,
+    AgentExecutionCompletedEvent,
+    TaskStartedEvent,
+    TaskCompletedEvent,
+    TaskFailedEvent,
+    FlowStartedEvent,
+    FlowFinishedEvent,
+    MethodExecutionStartedEvent,
+    MethodExecutionFinishedEvent,
+    MethodExecutionFailedEvent,
+    AgentExecutionErrorEvent,
+    ToolUsageFinishedEvent,
+    ToolUsageErrorEvent,
+    ToolUsageStartedEvent,
+]

src/crewai/utilities/events/flow_events.py

@@ -0,0 +1,71 @@
+from typing import Any, Dict, Optional, Union
+
+from pydantic import BaseModel
+
+from .base_events import CrewEvent
+
+
+class FlowEvent(CrewEvent):
+    """Base class for all flow events"""
+
+    type: str
+    flow_name: str
+
+
+class FlowStartedEvent(FlowEvent):
+    """Event emitted when a flow starts execution"""
+
+    flow_name: str
+    inputs: Optional[Dict[str, Any]] = None
+    type: str = "flow_started"
+
+
+class FlowCreatedEvent(FlowEvent):
+    """Event emitted when a flow is created"""
+
+    flow_name: str
+    type: str = "flow_created"
+
+
+class MethodExecutionStartedEvent(FlowEvent):
+    """Event emitted when a flow method starts execution"""
+
+    flow_name: str
+    method_name: str
+    state: Union[Dict[str, Any], BaseModel]
+    params: Optional[Dict[str, Any]] = None
+    type: str = "method_execution_started"
+
+
+class MethodExecutionFinishedEvent(FlowEvent):
+    """Event emitted when a flow method completes execution"""
+
+    flow_name: str
+    method_name: str
+    result: Any = None
+    state: Union[Dict[str, Any], BaseModel]
+    type: str = "method_execution_finished"
+
+
+class MethodExecutionFailedEvent(FlowEvent):
+    """Event emitted when a flow method fails execution"""
+
+    flow_name: str
+    method_name: str
+    error: Any
+    type: str = "method_execution_failed"
+
+
+class FlowFinishedEvent(FlowEvent):
+    """Event emitted when a flow completes execution"""
+
+    flow_name: str
+    result: Optional[Any] = None
+    type: str = "flow_finished"
+
+
+class FlowPlotEvent(FlowEvent):
+    """Event emitted when a flow plot is created"""
+
+    flow_name: str
+    type: str = "flow_plot"

src/crewai/utilities/events/task_events.py

@@ -0,0 +1,32 @@
+from typing import Any, Optional
+
+from crewai.tasks.task_output import TaskOutput
+from crewai.utilities.events.base_events import CrewEvent
+
+
+class TaskStartedEvent(CrewEvent):
+    """Event emitted when a task starts"""
+
+    type: str = "task_started"
+    context: Optional[str]
+
+
+class TaskCompletedEvent(CrewEvent):
+    """Event emitted when a task completes"""
+
+    output: TaskOutput
+    type: str = "task_completed"
+
+
+class TaskFailedEvent(CrewEvent):
+    """Event emitted when a task fails"""
+
+    error: str
+    type: str = "task_failed"
+
+
+class TaskEvaluationEvent(CrewEvent):
+    """Event emitted when a task evaluation is completed"""
+
+    type: str = "task_evaluation"
+    evaluation_type: str

src/crewai/utilities/events/third_party/__init__.py

@@ -0,0 +1 @@
+from .agentops_listener import agentops_listener

src/crewai/utilities/events/third_party/agentops_listener.py

@@ -0,0 +1,67 @@
+from typing import Optional
+
+from crewai.utilities.events import (
+    CrewKickoffCompletedEvent,
+    ToolUsageErrorEvent,
+    ToolUsageStartedEvent,
+)
+from crewai.utilities.events.base_event_listener import BaseEventListener
+from crewai.utilities.events.crew_events import CrewKickoffStartedEvent
+from crewai.utilities.events.task_events import TaskEvaluationEvent
+
+try:
+    import agentops
+
+    AGENTOPS_INSTALLED = True
+except ImportError:
+    AGENTOPS_INSTALLED = False
+
+
+class AgentOpsListener(BaseEventListener):
+    tool_event: Optional["agentops.ToolEvent"] = None
+    session: Optional["agentops.Session"] = None
+
+    def __init__(self):
+        super().__init__()
+
+    def setup_listeners(self, crewai_event_bus):
+        if not AGENTOPS_INSTALLED:
+            return
+
+        @crewai_event_bus.on(CrewKickoffStartedEvent)
+        def on_crew_kickoff_started(source, event: CrewKickoffStartedEvent):
+            self.session = agentops.init()
+            for agent in source.agents:
+                if self.session:
+                    self.session.create_agent(
+                        name=agent.role,
+                        agent_id=str(agent.id),
+                    )
+
+        @crewai_event_bus.on(CrewKickoffCompletedEvent)
+        def on_crew_kickoff_completed(source, event: CrewKickoffCompletedEvent):
+            if self.session:
+                self.session.end_session(
+                    end_state="Success",
+                    end_state_reason="Finished Execution",
+                )
+
+        @crewai_event_bus.on(ToolUsageStartedEvent)
+        def on_tool_usage_started(source, event: ToolUsageStartedEvent):
+            self.tool_event = agentops.ToolEvent(name=event.tool_name)
+            if self.session:
+                self.session.record(self.tool_event)
+
+        @crewai_event_bus.on(ToolUsageErrorEvent)
+        def on_tool_usage_error(source, event: ToolUsageErrorEvent):
+            agentops.ErrorEvent(exception=event.error, trigger_event=self.tool_event)
+
+        @crewai_event_bus.on(TaskEvaluationEvent)
+        def on_task_evaluation(source, event: TaskEvaluationEvent):
+            if self.session:
+                self.session.create_agent(
+                    name="Task Evaluator", agent_id=str(source.original_agent.id)
+                )
+
+
+agentops_listener = AgentOpsListener()

src/crewai/utilities/events/tool_usage_events.py

@@ -0,0 +1,64 @@
+from datetime import datetime
+from typing import Any, Callable, Dict
+
+from .base_events import CrewEvent
+
+
+class ToolUsageEvent(CrewEvent):
+    """Base event for tool usage tracking"""
+
+    agent_key: str
+    agent_role: str
+    tool_name: str
+    tool_args: Dict[str, Any] | str
+    tool_class: str
+    run_attempts: int | None = None
+    delegations: int | None = None
+
+    model_config = {"arbitrary_types_allowed": True}
+
+
+class ToolUsageStartedEvent(ToolUsageEvent):
+    """Event emitted when a tool execution is started"""
+
+    type: str = "tool_usage_started"
+
+
+class ToolUsageFinishedEvent(ToolUsageEvent):
+    """Event emitted when a tool execution is completed"""
+
+    started_at: datetime
+    finished_at: datetime
+    from_cache: bool = False
+    type: str = "tool_usage_finished"
+
+
+class ToolUsageErrorEvent(ToolUsageEvent):
+    """Event emitted when a tool execution encounters an error"""
+
+    error: Any
+    type: str = "tool_usage_error"
+
+
+class ToolValidateInputErrorEvent(ToolUsageEvent):
+    """Event emitted when a tool input validation encounters an error"""
+
+    error: Any
+    type: str = "tool_validate_input_error"
+
+
+class ToolSelectionErrorEvent(ToolUsageEvent):
+    """Event emitted when a tool selection encounters an error"""
+
+    error: Any
+    type: str = "tool_selection_error"
+
+
+class ToolExecutionErrorEvent(CrewEvent):
+    """Event emitted when a tool execution encounters an error"""
+
+    error: Any
+    type: str = "tool_execution_error"
+    tool_name: str
+    tool_args: Dict[str, Any]
+    tool_class: Callable

src/crewai/utilities/logger.py

@@ -8,8 +8,11 @@ from crewai.utilities.printer import Printer
 class Logger(BaseModel):
     verbose: bool = Field(default=False)
     _printer: Printer = PrivateAttr(default_factory=Printer)
+    default_color: str = Field(default="bold_yellow")
 
-    def log(self, level, message, color="bold_yellow"):
+    def log(self, level, message, color=None):
+        if color is None:
+            color = self.default_color
         if self.verbose:
             timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
             self._printer.print(

src/crewai/utilities/protocols.py

@@ -0,0 +1,12 @@
+from typing import Any, Protocol, runtime_checkable
+
+
+@runtime_checkable
+class AgentExecutorProtocol(Protocol):
+    """Protocol defining the expected interface for an agent executor."""
+
+    @property
+    def agent(self) -> Any: ...
+
+    @property
+    def task(self) -> Any: ...

tests/agent_test.py

@@ -1,6 +1,7 @@
 """Test Agent creation and execution basic functionality."""
 
 import os
+from datetime import datetime, timezone
 from unittest import mock
 from unittest.mock import patch
 
@@ -16,9 +17,10 @@ from crewai.llm import LLM
 from crewai.tools import tool
 from crewai.tools.tool_calling import InstructorToolCalling
 from crewai.tools.tool_usage import ToolUsage
-from crewai.tools.tool_usage_events import ToolUsageFinished
 from crewai.utilities import RPMController
-from crewai.utilities.events import Emitter
+from crewai.utilities.datetime_compat import UTC
+from crewai.utilities.events import crewai_event_bus
+from crewai.utilities.events.tool_usage_events import ToolUsageFinishedEvent
 
 
 def test_agent_llm_creation_with_env_vars():
@@ -154,15 +156,19 @@ def test_agent_execution_with_tools():
         agent=agent,
         expected_output="The result of the multiplication.",
     )
-    with patch.object(Emitter, "emit") as emit:
-        output = agent.execute_task(task)
-        assert output == "The result of the multiplication is 12."
-        assert emit.call_count == 1
-        args, _ = emit.call_args
-        assert isinstance(args[1], ToolUsageFinished)
-        assert not args[1].from_cache
-        assert args[1].tool_name == "multiplier"
-        assert args[1].tool_args == {"first_number": 3, "second_number": 4}
+    received_events = []
+
+    @crewai_event_bus.on(ToolUsageFinishedEvent)
+    def handle_tool_end(source, event):
+        received_events.append(event)
+
+    output = agent.execute_task(task)
+    assert output == "The result of the multiplication is 12."
+
+    assert len(received_events) == 1
+    assert isinstance(received_events[0], ToolUsageFinishedEvent)
+    assert received_events[0].tool_name == "multiplier"
+    assert received_events[0].tool_args == {"first_number": 3, "second_number": 4}
 
 
 @pytest.mark.vcr(filter_headers=["authorization"])
@@ -249,10 +255,14 @@ def test_cache_hitting():
         "multiplier-{'first_number': 3, 'second_number': 3}": 9,
         "multiplier-{'first_number': 12, 'second_number': 3}": 36,
     }
+    received_events = []
+
+    @crewai_event_bus.on(ToolUsageFinishedEvent)
+    def handle_tool_end(source, event):
+        received_events.append(event)
 
     with (
         patch.object(CacheHandler, "read") as read,
-        patch.object(Emitter, "emit") as emit,
     ):
         read.return_value = "0"
         task = Task(
@@ -265,10 +275,9 @@ def test_cache_hitting():
         read.assert_called_with(
             tool="multiplier", input={"first_number": 2, "second_number": 6}
         )
-        assert emit.call_count == 1
-        args, _ = emit.call_args
-        assert isinstance(args[1], ToolUsageFinished)
-        assert args[1].from_cache
+        assert len(received_events) == 1
+        assert isinstance(received_events[0], ToolUsageFinishedEvent)
+        assert received_events[0].from_cache
 
 
 @pytest.mark.vcr(filter_headers=["authorization"])
@@ -908,7 +917,10 @@ def test_tool_result_as_answer_is_the_final_answer_for_the_agent():
 
 @pytest.mark.vcr(filter_headers=["authorization"])
 def test_tool_usage_information_is_appended_to_agent():
+    from datetime import datetime, timezone
+
     from crewai.tools import BaseTool
+    from crewai.utilities.datetime_compat import UTC
 
     class MyCustomTool(BaseTool):
         name: str = "Decide Greetings"
@@ -917,30 +929,37 @@ def test_tool_usage_information_is_appended_to_agent():
         def _run(self) -> str:
             return "Howdy!"
 
-    agent1 = Agent(
-        role="Friendly Neighbor",
-        goal="Make everyone feel welcome",
-        backstory="You are the friendly neighbor",
-        tools=[MyCustomTool(result_as_answer=True)],
-    )
+    fixed_datetime = datetime(2025, 2, 10, 12, 0, 0, tzinfo=UTC)
+    with patch("crewai.tools.tool_usage.datetime", autospec=True) as mock_datetime:
+        mock_datetime.now.return_value = fixed_datetime
+        mock_datetime.fromtimestamp = datetime.fromtimestamp
+        mock_datetime.side_effect = lambda *args, **kw: datetime(*args, **kw)
 
-    greeting = Task(
-        description="Say an appropriate greeting.",
-        expected_output="The greeting.",
-        agent=agent1,
-    )
-    tasks = [greeting]
-    crew = Crew(agents=[agent1], tasks=tasks)
+        agent1 = Agent(
+            role="Friendly Neighbor",
+            goal="Make everyone feel welcome",
+            backstory="You are the friendly neighbor",
+            tools=[MyCustomTool(result_as_answer=True)],
+        )
 
-    crew.kickoff()
-    assert agent1.tools_results == [
-        {
-            "result": "Howdy!",
-            "tool_name": "Decide Greetings",
-            "tool_args": {},
-            "result_as_answer": True,
-        }
-    ]
+        greeting = Task(
+            description="Say an appropriate greeting.",
+            expected_output="The greeting.",
+            agent=agent1,
+        )
+        tasks = [greeting]
+        crew = Crew(agents=[agent1], tasks=tasks)
+
+        crew.kickoff()
+        assert agent1.tools_results == [
+            {
+                "result": "Howdy!",
+                "tool_name": "Decide Greetings",
+                "tool_args": {},
+                "result_as_answer": True,
+                "start_time": fixed_datetime,
+            }
+        ]
 
 
 def test_agent_definition_based_on_dict():

tests/cassettes/test_tool_execution_error_event.yaml

@@ -0,0 +1,112 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "user", "content": "Use the failing tool"}], "model":
+      "gpt-4o-mini", "stop": [], "tools": [{"type": "function", "function": {"name":
+      "failing_tool", "description": "This tool always fails.", "parameters": {"type":
+      "object", "properties": {"param": {"type": "string", "description": "A test
+      parameter"}}, "required": ["param"]}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '353'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-B2P4zoJZuES7Aom8ugEq1modz5Vsl\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739912761,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_F6fJxISpMKUBIGV6dd2vjRNG\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"failing_tool\",\n
+      \             \"arguments\": \"{\\\"param\\\":\\\"test\\\"}\"\n            }\n
+      \         }\n        ],\n        \"refusal\": null\n      },\n      \"logprobs\":
+      null,\n      \"finish_reason\": \"tool_calls\"\n    }\n  ],\n  \"usage\": {\n
+      \   \"prompt_tokens\": 51,\n    \"completion_tokens\": 15,\n    \"total_tokens\":
+      66,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\": 0,\n      \"audio_tokens\":
+      0\n    },\n    \"completion_tokens_details\": {\n      \"reasoning_tokens\":
+      0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\": 0,\n      \"rejected_prediction_tokens\":
+      0\n    }\n  },\n  \"service_tier\": \"default\",\n  \"system_fingerprint\":
+      \"fp_00428b782a\"\n}\n"
+    headers:
+      CF-RAY:
+      - 9140fa827f38eb1e-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 18 Feb 2025 21:06:02 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=xbuu3IQpCMh.43ZrqL1TRMECOc6QldgHV0hzOX1GrWI-1739912762-1.0.1.1-t7iyq5xMioPrwfeaHLvPT9rwRPp7Q9A9uIm69icH9dPxRD4xMA3cWqb1aXj1_e2IyAEQQWFe1UWjlmJ22aHh3Q;
+        path=/; expires=Tue, 18-Feb-25 21:36:02 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=x9l.Rhja8_wXDN.j8qcEU1PvvEqAwZp4Fd3s_aj4qwM-1739912762161-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '861'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999978'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_8666ec3aa6677cb346ba00993556051d
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/crew_test.py

@@ -6,7 +6,6 @@ from concurrent.futures import Future
 from unittest import mock
 from unittest.mock import MagicMock, patch
 
-import instructor
 import pydantic_core
 import pytest
 
@@ -15,15 +14,24 @@ from crewai.agents.cache import CacheHandler
 from crewai.crew import Crew
 from crewai.crews.crew_output import CrewOutput
 from crewai.knowledge.source.string_knowledge_source import StringKnowledgeSource
+from crewai.llm import LLM
 from crewai.memory.contextual.contextual_memory import ContextualMemory
 from crewai.process import Process
-from crewai.project import crew
 from crewai.task import Task
 from crewai.tasks.conditional_task import ConditionalTask
 from crewai.tasks.output_format import OutputFormat
 from crewai.tasks.task_output import TaskOutput
 from crewai.types.usage_metrics import UsageMetrics
 from crewai.utilities import Logger
+from crewai.utilities.events import (
+    CrewTrainCompletedEvent,
+    CrewTrainStartedEvent,
+    crewai_event_bus,
+)
+from crewai.utilities.events.crew_events import (
+    CrewTestCompletedEvent,
+    CrewTestStartedEvent,
+)
 from crewai.utilities.rpm_controller import RPMController
 from crewai.utilities.task_output_storage_handler import TaskOutputStorageHandler
 
@@ -843,8 +851,21 @@ def test_crew_verbose_output(capsys):
     crew.verbose = False
     crew._logger = Logger(verbose=False)
     crew.kickoff()
+    expected_listener_logs = [
+        "[🚀 CREW 'CREW' STARTED]",
+        "[📋 TASK STARTED: RESEARCH AI ADVANCEMENTS.]",
+        "[🤖 AGENT 'RESEARCHER' STARTED TASK]",
+        "[✅ AGENT 'RESEARCHER' COMPLETED TASK]",
+        "[✅ TASK COMPLETED: RESEARCH AI ADVANCEMENTS.]",
+        "[📋 TASK STARTED: WRITE ABOUT AI IN HEALTHCARE.]",
+        "[🤖 AGENT 'SENIOR WRITER' STARTED TASK]",
+        "[✅ AGENT 'SENIOR WRITER' COMPLETED TASK]",
+        "[✅ TASK COMPLETED: WRITE ABOUT AI IN HEALTHCARE.]",
+        "[✅ CREW 'CREW' COMPLETED]",
+    ]
     captured = capsys.readouterr()
-    assert captured.out == ""
+    for log in expected_listener_logs:
+        assert log in captured.out
 
 
 @pytest.mark.vcr(filter_headers=["authorization"])
@@ -1282,9 +1303,9 @@ def test_kickoff_for_each_invalid_input():
 
     crew = Crew(agents=[agent], tasks=[task])
 
-    with pytest.raises(TypeError):
+    with pytest.raises(pydantic_core._pydantic_core.ValidationError):
         # Pass a string instead of a list
-        crew.kickoff_for_each("invalid input")
+        crew.kickoff_for_each(["invalid input"])
 
 
 def test_kickoff_for_each_error_handling():
@@ -2568,6 +2589,16 @@ def test_crew_train_success(
     # Create a mock for the copied crew
     copy_mock.return_value = crew
 
+    received_events = []
+
+    @crewai_event_bus.on(CrewTrainStartedEvent)
+    def on_crew_train_started(source, event: CrewTrainStartedEvent):
+        received_events.append(event)
+
+    @crewai_event_bus.on(CrewTrainCompletedEvent)
+    def on_crew_train_completed(source, event: CrewTrainCompletedEvent):
+        received_events.append(event)
+
     crew.train(
         n_iterations=2, inputs={"topic": "AI"}, filename="trained_agents_data.pkl"
     )
@@ -2613,6 +2644,10 @@ def test_crew_train_success(
         ]
     )
 
+    assert len(received_events) == 2
+    assert isinstance(received_events[0], CrewTrainStartedEvent)
+    assert isinstance(received_events[1], CrewTrainCompletedEvent)
+
 
 def test_crew_train_error():
     task = Task(
@@ -3341,7 +3376,19 @@ def test_crew_testing_function(kickoff_mock, copy_mock, crew_evaluator):
     copy_mock.return_value = crew
 
     n_iterations = 2
-    crew.test(n_iterations, openai_model_name="gpt-4o-mini", inputs={"topic": "AI"})
+    llm_instance = LLM("gpt-4o-mini")
+
+    received_events = []
+
+    @crewai_event_bus.on(CrewTestStartedEvent)
+    def on_crew_test_started(source, event: CrewTestStartedEvent):
+        received_events.append(event)
+
+    @crewai_event_bus.on(CrewTestCompletedEvent)
+    def on_crew_test_completed(source, event: CrewTestCompletedEvent):
+        received_events.append(event)
+
+    crew.test(n_iterations, llm_instance, inputs={"topic": "AI"})
 
     # Ensure kickoff is called on the copied crew
     kickoff_mock.assert_has_calls(
@@ -3350,13 +3397,17 @@ def test_crew_testing_function(kickoff_mock, copy_mock, crew_evaluator):
 
     crew_evaluator.assert_has_calls(
         [
-            mock.call(crew, "gpt-4o-mini"),
+            mock.call(crew, llm_instance),
             mock.call().set_iteration(1),
             mock.call().set_iteration(2),
             mock.call().print_crew_evaluation_result(),
         ]
     )
 
+    assert len(received_events) == 2
+    assert isinstance(received_events[0], CrewTestStartedEvent)
+    assert isinstance(received_events[1], CrewTestCompletedEvent)
+
 
 @pytest.mark.vcr(filter_headers=["authorization"])
 def test_hierarchical_verbose_manager_agent():

tests/flow/test_state_utils.py

@@ -0,0 +1,150 @@
+from datetime import date, datetime
+from typing import List
+from unittest.mock import Mock
+
+import pytest
+from pydantic import BaseModel
+
+from crewai.flow import Flow
+from crewai.flow.state_utils import export_state, to_string
+
+
+class Address(BaseModel):
+    street: str
+    city: str
+    country: str
+
+
+class Person(BaseModel):
+    name: str
+    age: int
+    address: Address
+    birthday: date
+    skills: List[str]
+
+
+@pytest.fixture
+def mock_flow():
+    def create_flow(state):
+        flow = Mock(spec=Flow)
+        flow._state = state
+        return flow
+
+    return create_flow
+
+
+@pytest.mark.parametrize(
+    "test_input,expected",
+    [
+        ({"text": "hello world"}, {"text": "hello world"}),
+        ({"number": 42}, {"number": 42}),
+        ({"decimal": 3.14}, {"decimal": 3.14}),
+        ({"flag": True}, {"flag": True}),
+        ({"empty": None}, {"empty": None}),
+        ({"list": [1, 2, 3]}, {"list": [1, 2, 3]}),
+        ({"tuple": (1, 2, 3)}, {"tuple": [1, 2, 3]}),
+        ({"set": {1, 2, 3}}, {"set": [1, 2, 3]}),
+        ({"nested": [1, [2, 3], {4, 5}]}, {"nested": [1, [2, 3], [4, 5]]}),
+    ],
+)
+def test_basic_serialization(mock_flow, test_input, expected):
+    flow = mock_flow(test_input)
+    result = export_state(flow)
+    assert result == expected
+
+
+@pytest.mark.parametrize(
+    "input_date,expected",
+    [
+        (date(2024, 1, 1), "2024-01-01"),
+        (datetime(2024, 1, 1, 12, 30), "2024-01-01T12:30:00"),
+    ],
+)
+def test_temporal_serialization(mock_flow, input_date, expected):
+    flow = mock_flow({"date": input_date})
+    result = export_state(flow)
+    assert result["date"] == expected
+
+
+@pytest.mark.parametrize(
+    "key,value,expected_key_type",
+    [
+        (("tuple", "key"), "value", str),
+        (None, "value", str),
+        (123, "value", str),
+        ("normal", "value", str),
+    ],
+)
+def test_dictionary_key_serialization(mock_flow, key, value, expected_key_type):
+    flow = mock_flow({key: value})
+    result = export_state(flow)
+    assert len(result) == 1
+    result_key = next(iter(result.keys()))
+    assert isinstance(result_key, expected_key_type)
+    assert result[result_key] == value
+
+
+@pytest.mark.parametrize(
+    "callable_obj,expected_in_result",
+    [
+        (lambda x: x * 2, "lambda"),
+        (str.upper, "upper"),
+    ],
+)
+def test_callable_serialization(mock_flow, callable_obj, expected_in_result):
+    flow = mock_flow({"func": callable_obj})
+    result = export_state(flow)
+    assert isinstance(result["func"], str)
+    assert expected_in_result in result["func"].lower()
+
+
+def test_pydantic_model_serialization(mock_flow):
+    address = Address(street="123 Main St", city="Tech City", country="Pythonia")
+
+    person = Person(
+        name="John Doe",
+        age=30,
+        address=address,
+        birthday=date(1994, 1, 1),
+        skills=["Python", "Testing"],
+    )
+
+    flow = mock_flow(
+        {
+            "single_model": address,
+            "nested_model": person,
+            "model_list": [address, address],
+            "model_dict": {"home": address},
+        }
+    )
+
+    result = export_state(flow)
+    assert (
+        to_string(result)
+        == '{"single_model": {"street": "123 Main St", "city": "Tech City", "country": "Pythonia"}, "nested_model": {"name": "John Doe", "age": 30, "address": {"street": "123 Main St", "city": "Tech City", "country": "Pythonia"}, "birthday": "1994-01-01", "skills": ["Python", "Testing"]}, "model_list": [{"street": "123 Main St", "city": "Tech City", "country": "Pythonia"}, {"street": "123 Main St", "city": "Tech City", "country": "Pythonia"}], "model_dict": {"home": {"street": "123 Main St", "city": "Tech City", "country": "Pythonia"}}}'
+    )
+
+
+def test_depth_limit(mock_flow):
+    """Test max depth handling with a deeply nested structure"""
+
+    def create_nested(depth):
+        if depth == 0:
+            return "value"
+        return {"next": create_nested(depth - 1)}
+
+    deep_structure = create_nested(10)
+    flow = mock_flow(deep_structure)
+    result = export_state(flow)
+
+    assert result == {
+        "next": {
+            "next": {
+                "next": {
+                    "next": {
+                        "next": "{'next': {'next': {'next': {'next': {'next': 'value'}}}}}"
+                    }
+                }
+            }
+        }
+    }

tests/flow_test.py

@@ -7,12 +7,14 @@ import pytest
 from pydantic import BaseModel
 
 from crewai.flow.flow import Flow, and_, listen, or_, router, start
-from crewai.flow.flow_events import (
+from crewai.utilities.events import (
     FlowFinishedEvent,
     FlowStartedEvent,
     MethodExecutionFinishedEvent,
     MethodExecutionStartedEvent,
+    crewai_event_bus,
 )
+from crewai.utilities.events.flow_events import FlowPlotEvent
 
 
 def test_simple_sequential_flow():
@@ -434,90 +436,65 @@ def test_unstructured_flow_event_emission():
         @listen(finish_poem)
         def save_poem_to_database(self):
             # A method without args/kwargs to ensure events are sent correctly
-            pass
-
-    event_log = []
-
-    def handle_event(_, event):
-        event_log.append(event)
+            return "roses are red\nviolets are blue"
 
     flow = PoemFlow()
-    flow.event_emitter.connect(handle_event)
+    received_events = []
+
+    @crewai_event_bus.on(FlowStartedEvent)
+    def handle_flow_start(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(MethodExecutionStartedEvent)
+    def handle_method_start(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(FlowFinishedEvent)
+    def handle_flow_end(source, event):
+        received_events.append(event)
+
     flow.kickoff(inputs={"separator": ", "})
+    assert isinstance(received_events[0], FlowStartedEvent)
+    assert received_events[0].flow_name == "PoemFlow"
+    assert received_events[0].inputs == {"separator": ", "}
+    assert isinstance(received_events[0].timestamp, datetime)
 
-    assert isinstance(event_log[0], FlowStartedEvent)
-    assert event_log[0].flow_name == "PoemFlow"
-    assert event_log[0].inputs == {"separator": ", "}
-    assert isinstance(event_log[0].timestamp, datetime)
-
-    # Asserting for concurrent start method executions in a for loop as you
-    # can't guarantee ordering in asynchronous executions
-    for i in range(1, 5):
-        event = event_log[i]
+    # All subsequent events are MethodExecutionStartedEvent
+    for event in received_events[1:-1]:
+        assert isinstance(event, MethodExecutionStartedEvent)
+        assert event.flow_name == "PoemFlow"
         assert isinstance(event.state, dict)
         assert isinstance(event.state["id"], str)
+        assert event.state["separator"] == ", "
 
-        if event.method_name == "prepare_flower":
-            if isinstance(event, MethodExecutionStartedEvent):
-                assert event.params == {}
-                assert event.state["separator"] == ", "
-            elif isinstance(event, MethodExecutionFinishedEvent):
-                assert event.result == "foo"
-                assert event.state["flower"] == "roses"
-                assert event.state["separator"] == ", "
-            else:
-                assert False, "Unexpected event type for prepare_flower"
-        elif event.method_name == "prepare_color":
-            if isinstance(event, MethodExecutionStartedEvent):
-                assert event.params == {}
-                assert event.state["separator"] == ", "
-            elif isinstance(event, MethodExecutionFinishedEvent):
-                assert event.result == "bar"
-                assert event.state["color"] == "red"
-                assert event.state["separator"] == ", "
-            else:
-                assert False, "Unexpected event type for prepare_color"
-        else:
-            assert False, f"Unexpected method {event.method_name} in prepare events"
+    assert received_events[1].method_name == "prepare_flower"
+    assert received_events[1].params == {}
+    assert "flower" not in received_events[1].state
 
-    assert isinstance(event_log[5], MethodExecutionStartedEvent)
-    assert event_log[5].method_name == "write_first_sentence"
-    assert event_log[5].params == {}
-    assert isinstance(event_log[5].state, dict)
-    assert event_log[5].state["flower"] == "roses"
-    assert event_log[5].state["color"] == "red"
-    assert event_log[5].state["separator"] == ", "
+    assert received_events[2].method_name == "prepare_color"
+    assert received_events[2].params == {}
+    print("received_events[2]", received_events[2])
+    assert "flower" in received_events[2].state
 
-    assert isinstance(event_log[6], MethodExecutionFinishedEvent)
-    assert event_log[6].method_name == "write_first_sentence"
-    assert event_log[6].result == "roses are red"
+    assert received_events[3].method_name == "write_first_sentence"
+    assert received_events[3].params == {}
+    assert received_events[3].state["flower"] == "roses"
+    assert received_events[3].state["color"] == "red"
 
-    assert isinstance(event_log[7], MethodExecutionStartedEvent)
-    assert event_log[7].method_name == "finish_poem"
-    assert event_log[7].params == {"_0": "roses are red"}
-    assert isinstance(event_log[7].state, dict)
-    assert event_log[7].state["flower"] == "roses"
-    assert event_log[7].state["color"] == "red"
+    assert received_events[4].method_name == "finish_poem"
+    assert received_events[4].params == {"_0": "roses are red"}
+    assert received_events[4].state["flower"] == "roses"
+    assert received_events[4].state["color"] == "red"
 
-    assert isinstance(event_log[8], MethodExecutionFinishedEvent)
-    assert event_log[8].method_name == "finish_poem"
-    assert event_log[8].result == "roses are red, violets are blue"
+    assert received_events[5].method_name == "save_poem_to_database"
+    assert received_events[5].params == {}
+    assert received_events[5].state["flower"] == "roses"
+    assert received_events[5].state["color"] == "red"
 
-    assert isinstance(event_log[9], MethodExecutionStartedEvent)
-    assert event_log[9].method_name == "save_poem_to_database"
-    assert event_log[9].params == {}
-    assert isinstance(event_log[9].state, dict)
-    assert event_log[9].state["flower"] == "roses"
-    assert event_log[9].state["color"] == "red"
-
-    assert isinstance(event_log[10], MethodExecutionFinishedEvent)
-    assert event_log[10].method_name == "save_poem_to_database"
-    assert event_log[10].result is None
-
-    assert isinstance(event_log[11], FlowFinishedEvent)
-    assert event_log[11].flow_name == "PoemFlow"
-    assert event_log[11].result is None
-    assert isinstance(event_log[11].timestamp, datetime)
+    assert isinstance(received_events[6], FlowFinishedEvent)
+    assert received_events[6].flow_name == "PoemFlow"
+    assert received_events[6].result == "roses are red\nviolets are blue"
+    assert isinstance(received_events[6].timestamp, datetime)
 
 
 def test_structured_flow_event_emission():
@@ -538,40 +515,54 @@ def test_structured_flow_event_emission():
             self.state.sent = True
             return f"Welcome, {self.state.name}!"
 
-    event_log = []
-
-    def handle_event(_, event):
-        event_log.append(event)
-
     flow = OnboardingFlow()
-    flow.event_emitter.connect(handle_event)
     flow.kickoff(inputs={"name": "Anakin"})
 
-    assert isinstance(event_log[0], FlowStartedEvent)
-    assert event_log[0].flow_name == "OnboardingFlow"
-    assert event_log[0].inputs == {"name": "Anakin"}
-    assert isinstance(event_log[0].timestamp, datetime)
+    received_events = []
 
-    assert isinstance(event_log[1], MethodExecutionStartedEvent)
-    assert event_log[1].method_name == "user_signs_up"
+    @crewai_event_bus.on(FlowStartedEvent)
+    def handle_flow_start(source, event):
+        received_events.append(event)
 
-    assert isinstance(event_log[2], MethodExecutionFinishedEvent)
-    assert event_log[2].method_name == "user_signs_up"
+    @crewai_event_bus.on(MethodExecutionStartedEvent)
+    def handle_method_start(source, event):
+        received_events.append(event)
 
-    assert isinstance(event_log[3], MethodExecutionStartedEvent)
-    assert event_log[3].method_name == "send_welcome_message"
-    assert event_log[3].params == {}
-    assert getattr(event_log[3].state, "sent") is False
+    @crewai_event_bus.on(MethodExecutionFinishedEvent)
+    def handle_method_end(source, event):
+        received_events.append(event)
 
-    assert isinstance(event_log[4], MethodExecutionFinishedEvent)
-    assert event_log[4].method_name == "send_welcome_message"
-    assert getattr(event_log[4].state, "sent") is True
-    assert event_log[4].result == "Welcome, Anakin!"
+    @crewai_event_bus.on(FlowFinishedEvent)
+    def handle_flow_end(source, event):
+        received_events.append(event)
 
-    assert isinstance(event_log[5], FlowFinishedEvent)
-    assert event_log[5].flow_name == "OnboardingFlow"
-    assert event_log[5].result == "Welcome, Anakin!"
-    assert isinstance(event_log[5].timestamp, datetime)
+    flow.kickoff(inputs={"name": "Anakin"})
+
+    assert isinstance(received_events[0], FlowStartedEvent)
+    assert received_events[0].flow_name == "OnboardingFlow"
+    assert received_events[0].inputs == {"name": "Anakin"}
+    assert isinstance(received_events[0].timestamp, datetime)
+
+    assert isinstance(received_events[1], MethodExecutionStartedEvent)
+    assert received_events[1].method_name == "user_signs_up"
+
+    assert isinstance(received_events[2], MethodExecutionFinishedEvent)
+    assert received_events[2].method_name == "user_signs_up"
+
+    assert isinstance(received_events[3], MethodExecutionStartedEvent)
+    assert received_events[3].method_name == "send_welcome_message"
+    assert received_events[3].params == {}
+    assert getattr(received_events[3].state, "sent") is False
+
+    assert isinstance(received_events[4], MethodExecutionFinishedEvent)
+    assert received_events[4].method_name == "send_welcome_message"
+    assert getattr(received_events[4].state, "sent") is True
+    assert received_events[4].result == "Welcome, Anakin!"
+
+    assert isinstance(received_events[5], FlowFinishedEvent)
+    assert received_events[5].flow_name == "OnboardingFlow"
+    assert received_events[5].result == "Welcome, Anakin!"
+    assert isinstance(received_events[5].timestamp, datetime)
 
 
 def test_stateless_flow_event_emission():
@@ -593,30 +584,73 @@ def test_stateless_flow_event_emission():
         event_log.append(event)
 
     flow = StatelessFlow()
-    flow.event_emitter.connect(handle_event)
+    received_events = []
+
+    @crewai_event_bus.on(FlowStartedEvent)
+    def handle_flow_start(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(MethodExecutionStartedEvent)
+    def handle_method_start(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(MethodExecutionFinishedEvent)
+    def handle_method_end(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(FlowFinishedEvent)
+    def handle_flow_end(source, event):
+        received_events.append(event)
+
     flow.kickoff()
 
-    assert isinstance(event_log[0], FlowStartedEvent)
-    assert event_log[0].flow_name == "StatelessFlow"
-    assert event_log[0].inputs is None
-    assert isinstance(event_log[0].timestamp, datetime)
+    assert isinstance(received_events[0], FlowStartedEvent)
+    assert received_events[0].flow_name == "StatelessFlow"
+    assert received_events[0].inputs is None
+    assert isinstance(received_events[0].timestamp, datetime)
 
-    assert isinstance(event_log[1], MethodExecutionStartedEvent)
-    assert event_log[1].method_name == "init"
+    assert isinstance(received_events[1], MethodExecutionStartedEvent)
+    assert received_events[1].method_name == "init"
 
-    assert isinstance(event_log[2], MethodExecutionFinishedEvent)
-    assert event_log[2].method_name == "init"
+    assert isinstance(received_events[2], MethodExecutionFinishedEvent)
+    assert received_events[2].method_name == "init"
 
-    assert isinstance(event_log[3], MethodExecutionStartedEvent)
-    assert event_log[3].method_name == "process"
+    assert isinstance(received_events[3], MethodExecutionStartedEvent)
+    assert received_events[3].method_name == "process"
 
-    assert isinstance(event_log[4], MethodExecutionFinishedEvent)
-    assert event_log[4].method_name == "process"
+    assert isinstance(received_events[4], MethodExecutionFinishedEvent)
+    assert received_events[4].method_name == "process"
 
-    assert isinstance(event_log[5], FlowFinishedEvent)
-    assert event_log[5].flow_name == "StatelessFlow"
+    assert isinstance(received_events[5], FlowFinishedEvent)
+    assert received_events[5].flow_name == "StatelessFlow"
     assert (
-        event_log[5].result
+        received_events[5].result
         == "Deeds will not be less valiant because they are unpraised."
     )
-    assert isinstance(event_log[5].timestamp, datetime)
+    assert isinstance(received_events[5].timestamp, datetime)
+
+
+def test_flow_plotting():
+    class StatelessFlow(Flow):
+        @start()
+        def init(self):
+            return "Initializing flow..."
+
+        @listen(init)
+        def process(self):
+            return "Deeds will not be less valiant because they are unpraised."
+
+    flow = StatelessFlow()
+    flow.kickoff()
+    received_events = []
+
+    @crewai_event_bus.on(FlowPlotEvent)
+    def handle_flow_plot(source, event):
+        received_events.append(event)
+
+    flow.plot("test_flow")
+
+    assert len(received_events) == 1
+    assert isinstance(received_events[0], FlowPlotEvent)
+    assert received_events[0].flow_name == "StatelessFlow"
+    assert isinstance(received_events[0].timestamp, datetime)

tests/llm_test.py

@@ -7,7 +7,8 @@ from pydantic import BaseModel
 
 from crewai.agents.agent_builder.utilities.base_token_process import TokenProcess
 from crewai.llm import LLM
-from crewai.tools import tool
+from crewai.utilities.events import crewai_event_bus
+from crewai.utilities.events.tool_usage_events import ToolExecutionErrorEvent
 from crewai.utilities.token_counter_callback import TokenCalcHandler
 
 
@@ -291,32 +292,36 @@ def anthropic_llm():
     """Fixture providing an Anthropic LLM instance."""
     return LLM(model="anthropic/claude-3-sonnet")
 
+
 @pytest.fixture
 def system_message():
     """Fixture providing a system message."""
     return {"role": "system", "content": "test"}
 
+
 @pytest.fixture
 def user_message():
     """Fixture providing a user message."""
     return {"role": "user", "content": "test"}
 
+
 def test_anthropic_message_formatting_edge_cases(anthropic_llm):
     """Test edge cases for Anthropic message formatting."""
     # Test None messages
     with pytest.raises(TypeError, match="Messages cannot be None"):
         anthropic_llm._format_messages_for_provider(None)
-        
+
     # Test empty message list
     formatted = anthropic_llm._format_messages_for_provider([])
     assert len(formatted) == 1
     assert formatted[0]["role"] == "user"
     assert formatted[0]["content"] == "."
-    
+
     # Test invalid message format
     with pytest.raises(TypeError, match="Invalid message format"):
         anthropic_llm._format_messages_for_provider([{"invalid": "message"}])
 
+
 def test_anthropic_model_detection():
     """Test Anthropic model detection with various formats."""
     models = [
@@ -327,11 +332,12 @@ def test_anthropic_model_detection():
         ("", False),
         ("anthropomorphic", False),  # Should not match partial words
     ]
-    
+
     for model, expected in models:
         llm = LLM(model=model)
         assert llm.is_anthropic == expected, f"Failed for model: {model}"
 
+
 def test_anthropic_message_formatting(anthropic_llm, system_message, user_message):
     """Test Anthropic message formatting with fixtures."""
     # Test when first message is system
@@ -371,3 +377,51 @@ def test_deepseek_r1_with_open_router():
     result = llm.call("What is the capital of France?")
     assert isinstance(result, str)
     assert "Paris" in result
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_tool_execution_error_event():
+    llm = LLM(model="gpt-4o-mini")
+
+    def failing_tool(param: str) -> str:
+        """This tool always fails."""
+        raise Exception("Tool execution failed!")
+
+    tool_schema = {
+        "type": "function",
+        "function": {
+            "name": "failing_tool",
+            "description": "This tool always fails.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "param": {"type": "string", "description": "A test parameter"}
+                },
+                "required": ["param"],
+            },
+        },
+    }
+
+    received_events = []
+
+    @crewai_event_bus.on(ToolExecutionErrorEvent)
+    def event_handler(source, event):
+        received_events.append(event)
+
+    available_functions = {"failing_tool": failing_tool}
+
+    messages = [{"role": "user", "content": "Use the failing tool"}]
+
+    llm.call(
+        messages,
+        tools=[tool_schema],
+        available_functions=available_functions,
+    )
+
+    assert len(received_events) == 1
+    event = received_events[0]
+    assert isinstance(event, ToolExecutionErrorEvent)
+    assert event.tool_name == "failing_tool"
+    assert event.tool_args == {"param": "test"}
+    assert event.tool_class == failing_tool
+    assert "Tool execution failed!" in event.error

tests/test_flow_persistence.py

@@ -13,11 +13,12 @@ from crewai.flow.persistence.sqlite import SQLiteFlowPersistence
 
 class TestState(FlowState):
     """Test state model with required id field."""
+
     counter: int = 0
     message: str = ""
 
 
-def test_persist_decorator_saves_state(tmp_path):
+def test_persist_decorator_saves_state(tmp_path, caplog):
     """Test that @persist decorator saves state in SQLite."""
     db_path = os.path.join(tmp_path, "test_flows.db")
     persistence = SQLiteFlowPersistence(db_path)
@@ -73,7 +74,6 @@ def test_flow_state_restoration(tmp_path):
 
     # First flow execution to create initial state
     class RestorableFlow(Flow[TestState]):
-
         @start()
         @persist(persistence)
         def set_message(self):
@@ -89,10 +89,7 @@ def test_flow_state_restoration(tmp_path):
 
     # Test case 1: Restore using restore_uuid with field override
     flow2 = RestorableFlow(persistence=persistence)
-    flow2.kickoff(inputs={
-        "id": original_uuid,
-        "counter": 43
-    })
+    flow2.kickoff(inputs={"id": original_uuid, "counter": 43})
 
     # Verify state restoration and selective field override
     assert flow2.state.id == original_uuid
@@ -101,10 +98,7 @@ def test_flow_state_restoration(tmp_path):
 
     # Test case 2: Restore using kwargs['id']
     flow3 = RestorableFlow(persistence=persistence)
-    flow3.kickoff(inputs={
-        "id": original_uuid,
-        "message": "Updated message"
-    })
+    flow3.kickoff(inputs={"id": original_uuid, "message": "Updated message"})
 
     # Verify state restoration and selective field override
     assert flow3.state.id == original_uuid
@@ -174,3 +168,43 @@ def test_multiple_method_persistence(tmp_path):
     final_state = flow2.state
     assert final_state.counter == 99999
     assert final_state.message == "Step 99999"
+
+
+def test_persist_decorator_verbose_logging(tmp_path, caplog):
+    """Test that @persist decorator's verbose parameter controls logging."""
+    # Set logging level to ensure we capture all logs
+    caplog.set_level("INFO")
+
+    db_path = os.path.join(tmp_path, "test_flows.db")
+    persistence = SQLiteFlowPersistence(db_path)
+
+    # Test with verbose=False (default)
+    class QuietFlow(Flow[Dict[str, str]]):
+        initial_state = dict()
+
+        @start()
+        @persist(persistence)  # Default verbose=False
+        def init_step(self):
+            self.state["message"] = "Hello, World!"
+            self.state["id"] = "test-uuid-1"
+
+    flow = QuietFlow(persistence=persistence)
+    flow.kickoff()
+    assert "Saving flow state" not in caplog.text
+
+    # Clear the log
+    caplog.clear()
+
+    # Test with verbose=True
+    class VerboseFlow(Flow[Dict[str, str]]):
+        initial_state = dict()
+
+        @start()
+        @persist(persistence, verbose=True)
+        def init_step(self):
+            self.state["message"] = "Hello, World!"
+            self.state["id"] = "test-uuid-2"
+
+    flow = VerboseFlow(persistence=persistence)
+    flow.kickoff()
+    assert "Saving flow state" in caplog.text

tests/tools/test_tool_usage.py

@@ -1,6 +1,6 @@
 import json
 import random
-from unittest.mock import MagicMock
+from unittest.mock import MagicMock, patch
 
 import pytest
 from pydantic import BaseModel, Field
@@ -8,6 +8,11 @@ from pydantic import BaseModel, Field
 from crewai import Agent, Task
 from crewai.tools import BaseTool
 from crewai.tools.tool_usage import ToolUsage
+from crewai.utilities.events import crewai_event_bus
+from crewai.utilities.events.tool_usage_events import (
+    ToolSelectionErrorEvent,
+    ToolValidateInputErrorEvent,
+)
 
 
 class RandomNumberToolInput(BaseModel):
@@ -226,7 +231,7 @@ def test_validate_tool_input_with_special_characters():
     )
 
     # Input with special characters
-    tool_input = '{"message": "Hello, world! \u263A", "valid": True}'
+    tool_input = '{"message": "Hello, world! \u263a", "valid": True}'
     expected_arguments = {"message": "Hello, world! ☺", "valid": True}
 
     arguments = tool_usage._validate_tool_input(tool_input)
@@ -331,6 +336,19 @@ def test_validate_tool_input_with_trailing_commas():
 
 
 def test_validate_tool_input_invalid_input():
+    # Create mock agent with proper string values
+    mock_agent = MagicMock()
+    mock_agent.key = "test_agent_key"  # Must be a string
+    mock_agent.role = "test_agent_role"  # Must be a string
+    mock_agent._original_role = "test_agent_role"  # Must be a string
+    mock_agent.i18n = MagicMock()
+    mock_agent.verbose = False
+
+    # Create mock action with proper string value
+    mock_action = MagicMock()
+    mock_action.tool = "test_tool"  # Must be a string
+    mock_action.tool_input = "test_input"  # Must be a string
+
     tool_usage = ToolUsage(
         tools_handler=MagicMock(),
         tools=[],
@@ -339,8 +357,8 @@ def test_validate_tool_input_invalid_input():
         tools_names="",
         task=MagicMock(),
         function_calling_llm=None,
-        agent=MagicMock(),
-        action=MagicMock(),
+        agent=mock_agent,
+        action=mock_action,
     )
 
     invalid_inputs = [
@@ -360,7 +378,7 @@ def test_validate_tool_input_invalid_input():
 
     # Test for None input separately
     arguments = tool_usage._validate_tool_input(None)
-    assert arguments == {}  # Expecting an empty dictionary
+    assert arguments == {}
 
 
 def test_validate_tool_input_complex_structure():
@@ -468,18 +486,141 @@ def test_validate_tool_input_large_json_content():
     assert arguments == expected_arguments
 
 
-def test_validate_tool_input_none_input():
+def test_tool_selection_error_event_direct():
+    """Test tool selection error event emission directly from ToolUsage class."""
+    mock_agent = MagicMock()
+    mock_agent.key = "test_key"
+    mock_agent.role = "test_role"
+    mock_agent.i18n = MagicMock()
+    mock_agent.verbose = False
+
+    mock_task = MagicMock()
+    mock_tools_handler = MagicMock()
+
+    class TestTool(BaseTool):
+        name: str = "Test Tool"
+        description: str = "A test tool"
+
+        def _run(self, input: dict) -> str:
+            return "test result"
+
+    test_tool = TestTool()
+
     tool_usage = ToolUsage(
-        tools_handler=MagicMock(),
-        tools=[],
-        original_tools=[],
-        tools_description="",
-        tools_names="",
-        task=MagicMock(),
+        tools_handler=mock_tools_handler,
+        tools=[test_tool],
+        original_tools=[test_tool],
+        tools_description="Test Tool Description",
+        tools_names="Test Tool",
+        task=mock_task,
         function_calling_llm=None,
-        agent=MagicMock(),
+        agent=mock_agent,
         action=MagicMock(),
     )
 
-    arguments = tool_usage._validate_tool_input(None)
-    assert arguments == {}  # Expecting an empty dictionary
+    received_events = []
+
+    @crewai_event_bus.on(ToolSelectionErrorEvent)
+    def event_handler(source, event):
+        received_events.append(event)
+
+    with pytest.raises(Exception) as exc_info:
+        tool_usage._select_tool("Non Existent Tool")
+    assert len(received_events) == 1
+    event = received_events[0]
+    assert isinstance(event, ToolSelectionErrorEvent)
+    assert event.agent_key == "test_key"
+    assert event.agent_role == "test_role"
+    assert event.tool_name == "Non Existent Tool"
+    assert event.tool_args == {}
+    assert event.tool_class == "Test Tool Description"
+    assert "don't exist" in event.error
+
+    received_events.clear()
+    with pytest.raises(Exception) as exc_info:
+        tool_usage._select_tool("")
+
+    assert len(received_events) == 1
+    event = received_events[0]
+    assert isinstance(event, ToolSelectionErrorEvent)
+    assert event.agent_key == "test_key"
+    assert event.agent_role == "test_role"
+    assert event.tool_name == ""
+    assert event.tool_args == {}
+    assert event.tool_class == "Test Tool Description"
+    assert "forgot the Action name" in event.error
+
+
+def test_tool_validate_input_error_event():
+    """Test tool validation input error event emission from ToolUsage class."""
+    # Mock agent and required components
+    mock_agent = MagicMock()
+    mock_agent.key = "test_key"
+    mock_agent.role = "test_role"
+    mock_agent.verbose = False
+    mock_agent._original_role = "test_role"
+
+    # Mock i18n with error message
+    mock_i18n = MagicMock()
+    mock_i18n.errors.return_value = (
+        "Tool input must be a valid dictionary in JSON or Python literal format"
+    )
+    mock_agent.i18n = mock_i18n
+
+    # Mock task and tools handler
+    mock_task = MagicMock()
+    mock_tools_handler = MagicMock()
+
+    # Mock printer
+    mock_printer = MagicMock()
+
+    # Create test tool
+    class TestTool(BaseTool):
+        name: str = "Test Tool"
+        description: str = "A test tool"
+
+        def _run(self, input: dict) -> str:
+            return "test result"
+
+    test_tool = TestTool()
+
+    # Create ToolUsage instance
+    tool_usage = ToolUsage(
+        tools_handler=mock_tools_handler,
+        tools=[test_tool],
+        original_tools=[test_tool],
+        tools_description="Test Tool Description",
+        tools_names="Test Tool",
+        task=mock_task,
+        function_calling_llm=None,
+        agent=mock_agent,
+        action=MagicMock(tool="test_tool"),
+    )
+    tool_usage._printer = mock_printer
+
+    # Mock all parsing attempts to fail
+    with (
+        patch("json.loads", side_effect=json.JSONDecodeError("Test Error", "", 0)),
+        patch("ast.literal_eval", side_effect=ValueError),
+        patch("json5.loads", side_effect=json.JSONDecodeError("Test Error", "", 0)),
+        patch("json_repair.repair_json", side_effect=Exception("Failed to repair")),
+    ):
+        received_events = []
+
+        @crewai_event_bus.on(ToolValidateInputErrorEvent)
+        def event_handler(source, event):
+            received_events.append(event)
+
+        # Test invalid input
+        invalid_input = "invalid json {[}"
+        with pytest.raises(Exception) as exc_info:
+            tool_usage._validate_tool_input(invalid_input)
+
+        # Verify event was emitted
+        assert len(received_events) == 1, "Expected one event to be emitted"
+        event = received_events[0]
+        assert isinstance(event, ToolValidateInputErrorEvent)
+        assert event.agent_key == "test_key"
+        assert event.agent_role == "test_role"
+        assert event.tool_name == "test_tool"
+        assert "must be a valid dictionary" in event.error

tests/traces/test_unified_trace_controller.py

@@ -0,0 +1,361 @@
+import os
+from datetime import datetime
+from unittest.mock import MagicMock, patch
+from uuid import UUID
+
+import pytest
+
+from crewai.traces.context import TraceContext
+from crewai.traces.enums import CrewType, RunType, TraceType
+from crewai.traces.models import (
+    CrewTrace,
+    FlowStepIO,
+    LLMRequest,
+    LLMResponse,
+)
+from crewai.traces.unified_trace_controller import (
+    UnifiedTraceController,
+    init_crew_main_trace,
+    init_flow_main_trace,
+    should_trace,
+    trace_flow_step,
+    trace_llm_call,
+)
+from crewai.utilities.datetime_compat import UTC
+
+
+class TestUnifiedTraceController:
+    @pytest.fixture
+    def basic_trace_controller(self):
+        return UnifiedTraceController(
+            trace_type=TraceType.LLM_CALL,
+            run_type=RunType.KICKOFF,
+            crew_type=CrewType.CREW,
+            run_id="test-run-id",
+            agent_role="test-agent",
+            task_name="test-task",
+            task_description="test description",
+            task_id="test-task-id",
+        )
+
+    def test_initialization(self, basic_trace_controller):
+        """Test basic initialization of UnifiedTraceController"""
+        assert basic_trace_controller.trace_type == TraceType.LLM_CALL
+        assert basic_trace_controller.run_type == RunType.KICKOFF
+        assert basic_trace_controller.crew_type == CrewType.CREW
+        assert basic_trace_controller.run_id == "test-run-id"
+        assert basic_trace_controller.agent_role == "test-agent"
+        assert basic_trace_controller.task_name == "test-task"
+        assert basic_trace_controller.task_description == "test description"
+        assert basic_trace_controller.task_id == "test-task-id"
+        assert basic_trace_controller.status == "running"
+        assert isinstance(UUID(basic_trace_controller.trace_id), UUID)
+
+    def test_start_trace(self, basic_trace_controller):
+        """Test starting a trace"""
+        result = basic_trace_controller.start_trace()
+        assert result == basic_trace_controller
+        assert basic_trace_controller.start_time is not None
+        assert isinstance(basic_trace_controller.start_time, datetime)
+
+    def test_end_trace_success(self, basic_trace_controller):
+        """Test ending a trace successfully"""
+        basic_trace_controller.start_trace()
+        basic_trace_controller.end_trace(result={"test": "result"})
+
+        assert basic_trace_controller.end_time is not None
+        assert basic_trace_controller.status == "completed"
+        assert basic_trace_controller.error is None
+        assert basic_trace_controller.context.get("response") == {"test": "result"}
+
+    def test_end_trace_with_error(self, basic_trace_controller):
+        """Test ending a trace with an error"""
+        basic_trace_controller.start_trace()
+        basic_trace_controller.end_trace(error="Test error occurred")
+
+        assert basic_trace_controller.end_time is not None
+        assert basic_trace_controller.status == "error"
+        assert basic_trace_controller.error == "Test error occurred"
+
+    def test_add_child_trace(self, basic_trace_controller):
+        """Test adding a child trace"""
+        child_trace = {"id": "child-1", "type": "test"}
+        basic_trace_controller.add_child_trace(child_trace)
+        assert len(basic_trace_controller.children) == 1
+        assert basic_trace_controller.children[0] == child_trace
+
+    def test_to_crew_trace_llm_call(self):
+        """Test converting to CrewTrace for LLM call"""
+        test_messages = [{"role": "user", "content": "test"}]
+        test_response = {
+            "content": "test response",
+            "finish_reason": "stop",
+        }
+
+        controller = UnifiedTraceController(
+            trace_type=TraceType.LLM_CALL,
+            run_type=RunType.KICKOFF,
+            crew_type=CrewType.CREW,
+            run_id="test-run-id",
+            context={
+                "messages": test_messages,
+                "temperature": 0.7,
+                "max_tokens": 100,
+            },
+        )
+
+        # Set model and messages in the context
+        controller.context["model"] = "gpt-4"
+        controller.context["messages"] = test_messages
+
+        controller.start_trace()
+        controller.end_trace(result=test_response)
+
+        crew_trace = controller.to_crew_trace()
+        assert isinstance(crew_trace, CrewTrace)
+        assert isinstance(crew_trace.request, LLMRequest)
+        assert isinstance(crew_trace.response, LLMResponse)
+        assert crew_trace.request.model == "gpt-4"
+        assert crew_trace.request.messages == test_messages
+        assert crew_trace.response.content == test_response["content"]
+        assert crew_trace.response.finish_reason == test_response["finish_reason"]
+
+    def test_to_crew_trace_flow_step(self):
+        """Test converting to CrewTrace for flow step"""
+        flow_step_data = {
+            "function_name": "test_function",
+            "inputs": {"param1": "value1"},
+            "metadata": {"meta": "data"},
+        }
+
+        controller = UnifiedTraceController(
+            trace_type=TraceType.FLOW_STEP,
+            run_type=RunType.KICKOFF,
+            crew_type=CrewType.FLOW,
+            run_id="test-run-id",
+            flow_step=flow_step_data,
+        )
+
+        controller.start_trace()
+        controller.end_trace(result="test result")
+
+        crew_trace = controller.to_crew_trace()
+        assert isinstance(crew_trace, CrewTrace)
+        assert isinstance(crew_trace.flow_step, FlowStepIO)
+        assert crew_trace.flow_step.function_name == "test_function"
+        assert crew_trace.flow_step.inputs == {"param1": "value1"}
+        assert crew_trace.flow_step.outputs == {"result": "test result"}
+
+    def test_should_trace(self):
+        """Test should_trace function"""
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            assert should_trace() is True
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "false"}):
+            assert should_trace() is False
+
+        with patch.dict(os.environ, clear=True):
+            assert should_trace() is False
+
+    @pytest.mark.asyncio
+    async def test_trace_flow_step_decorator(self):
+        """Test trace_flow_step decorator"""
+
+        class TestFlow:
+            flow_id = "test-flow-id"
+
+            @trace_flow_step
+            async def test_method(self, method_name, method, *args, **kwargs):
+                return "test result"
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            flow = TestFlow()
+            result = await flow.test_method("test_method", lambda x: x, arg1="value1")
+            assert result == "test result"
+
+    def test_trace_llm_call_decorator(self):
+        """Test trace_llm_call decorator"""
+
+        class TestLLM:
+            model = "gpt-4"
+            temperature = 0.7
+            max_tokens = 100
+            stop = None
+
+            def _get_execution_context(self):
+                return MagicMock(), MagicMock()
+
+            def _get_new_messages(self, messages):
+                return messages
+
+            def _get_new_tool_results(self, agent):
+                return []
+
+            @trace_llm_call
+            def test_method(self, params):
+                return {
+                    "choices": [
+                        {
+                            "message": {"content": "test response"},
+                            "finish_reason": "stop",
+                        }
+                    ],
+                    "usage": {
+                        "total_tokens": 50,
+                        "prompt_tokens": 20,
+                        "completion_tokens": 30,
+                    },
+                }
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            llm = TestLLM()
+            result = llm.test_method({"messages": []})
+            assert result["choices"][0]["message"]["content"] == "test response"
+
+    def test_init_crew_main_trace_kickoff(self):
+        """Test init_crew_main_trace in kickoff mode"""
+        trace_context = None
+
+        class TestCrew:
+            id = "test-crew-id"
+            _test = False
+            _train = False
+
+        @init_crew_main_trace
+        def test_method(self):
+            nonlocal trace_context
+            trace_context = TraceContext.get_current()
+            return "test result"
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            crew = TestCrew()
+            result = test_method(crew)
+            assert result == "test result"
+            assert trace_context is not None
+            assert trace_context.trace_type == TraceType.LLM_CALL
+            assert trace_context.run_type == RunType.KICKOFF
+            assert trace_context.crew_type == CrewType.CREW
+            assert trace_context.run_id == str(crew.id)
+
+    def test_init_crew_main_trace_test_mode(self):
+        """Test init_crew_main_trace in test mode"""
+        trace_context = None
+
+        class TestCrew:
+            id = "test-crew-id"
+            _test = True
+            _train = False
+
+        @init_crew_main_trace
+        def test_method(self):
+            nonlocal trace_context
+            trace_context = TraceContext.get_current()
+            return "test result"
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            crew = TestCrew()
+            result = test_method(crew)
+            assert result == "test result"
+            assert trace_context is not None
+            assert trace_context.run_type == RunType.TEST
+
+    def test_init_crew_main_trace_train_mode(self):
+        """Test init_crew_main_trace in train mode"""
+        trace_context = None
+
+        class TestCrew:
+            id = "test-crew-id"
+            _test = False
+            _train = True
+
+        @init_crew_main_trace
+        def test_method(self):
+            nonlocal trace_context
+            trace_context = TraceContext.get_current()
+            return "test result"
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            crew = TestCrew()
+            result = test_method(crew)
+            assert result == "test result"
+            assert trace_context is not None
+            assert trace_context.run_type == RunType.TRAIN
+
+    @pytest.mark.asyncio
+    async def test_init_flow_main_trace(self):
+        """Test init_flow_main_trace decorator"""
+        trace_context = None
+        test_inputs = {"test": "input"}
+
+        class TestFlow:
+            flow_id = "test-flow-id"
+
+            @init_flow_main_trace
+            async def test_method(self, **kwargs):
+                nonlocal trace_context
+                trace_context = TraceContext.get_current()
+                # Verify the context is set during execution
+                assert trace_context.context["context"]["inputs"] == test_inputs
+                return "test result"
+
+        with patch.dict(os.environ, {"CREWAI_ENABLE_TRACING": "true"}):
+            flow = TestFlow()
+            result = await flow.test_method(inputs=test_inputs)
+            assert result == "test result"
+            assert trace_context is not None
+            assert trace_context.trace_type == TraceType.FLOW_STEP
+            assert trace_context.crew_type == CrewType.FLOW
+            assert trace_context.run_type == RunType.KICKOFF
+            assert trace_context.run_id == str(flow.flow_id)
+            assert trace_context.context["context"]["inputs"] == test_inputs
+
+    def test_trace_context_management(self):
+        """Test TraceContext management"""
+        trace1 = UnifiedTraceController(
+            trace_type=TraceType.LLM_CALL,
+            run_type=RunType.KICKOFF,
+            crew_type=CrewType.CREW,
+            run_id="test-run-1",
+        )
+
+        trace2 = UnifiedTraceController(
+            trace_type=TraceType.FLOW_STEP,
+            run_type=RunType.TEST,
+            crew_type=CrewType.FLOW,
+            run_id="test-run-2",
+        )
+
+        # Test that context is initially empty
+        assert TraceContext.get_current() is None
+
+        # Test setting and getting context
+        with TraceContext.set_current(trace1):
+            assert TraceContext.get_current() == trace1
+
+            # Test nested context
+            with TraceContext.set_current(trace2):
+                assert TraceContext.get_current() == trace2
+
+            # Test context restoration after nested block
+            assert TraceContext.get_current() == trace1
+
+        # Test context cleanup after with block
+        assert TraceContext.get_current() is None
+
+    def test_trace_context_error_handling(self):
+        """Test TraceContext error handling"""
+        trace = UnifiedTraceController(
+            trace_type=TraceType.LLM_CALL,
+            run_type=RunType.KICKOFF,
+            crew_type=CrewType.CREW,
+            run_id="test-run",
+        )
+
+        # Test that context is properly cleaned up even if an error occurs
+        try:
+            with TraceContext.set_current(trace):
+                raise ValueError("Test error")
+        except ValueError:
+            pass
+
+        assert TraceContext.get_current() is None

tests/utilities/cassettes/test_agent_emits_execution_started_and_completed_events.yaml

@@ -0,0 +1,243 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzTXAk4GatJOmLO9sEOCCITIjf1Dx\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739214900,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90fe6ce92eba67b3-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Mon, 10 Feb 2025 19:15:01 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=pjX1I6y8RlqCjS.gvOqvXk4vM69UNwFwmslh1BhALNg-1739214901-1.0.1.1-nJcNlSdNcug82eDl7KSvteLbsg0xCiEh2yI1TZX2jMAblL7AMQ8LFhvXkJLlAMfk49RMzRzWy2aiQgeM7WRHPg;
+        path=/; expires=Mon, 10-Feb-25 19:45:01 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=efIHP1NUsh1dFewGJBu4YoBu6hhGa8vjOOKQglYQGno-1739214901306-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '571'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_a95183a7a85e6bdfe381b2510bf70f34
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=pjX1I6y8RlqCjS.gvOqvXk4vM69UNwFwmslh1BhALNg-1739214901-1.0.1.1-nJcNlSdNcug82eDl7KSvteLbsg0xCiEh2yI1TZX2jMAblL7AMQ8LFhvXkJLlAMfk49RMzRzWy2aiQgeM7WRHPg;
+        _cfuvid=efIHP1NUsh1dFewGJBu4YoBu6hhGa8vjOOKQglYQGno-1739214901306-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzTXDcgKWq3yosIyBal8LcY8dDrn1\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739214903,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_c41SAnqyEKNXEAZd5XV3jKF3\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Consider specifying
+      the tone or context of the greeting for more engaging interactions.\\\",\\\"Clarify
+      if additional greetings or responses are acceptable to enhance the task's scope.\\\"],\\\"quality\\\":10,\\\"entities\\\":[]
+      }\"\n            }\n          }\n        ],\n        \"refusal\": null\n      },\n
+      \     \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n  ],\n
+      \ \"usage\": {\n    \"prompt_tokens\": 273,\n    \"completion_tokens\": 43,\n
+      \   \"total_tokens\": 316,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\":
+      0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\": {\n
+      \     \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 90fe6cf8c96e67b3-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Mon, 10 Feb 2025 19:15:04 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1181'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_b2286c8ae6f9b2a42f46a3e2c52b4211
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_end_kickoff_event.yaml

@@ -0,0 +1,315 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=4s6sWmJ49B9F_wNc1STtdZF1nikfl6uN9_ov3Xzfa8U-1738698987-1.0.1.1-lmbRRS1MHrDbnU93Gh16CP3qNczxxIrQnyBU7vpHSwNf6PdmuWOHKd1mkl5SBx6rg7p1NLaNUMyqDDcE0Mvjzw;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJK2OCJSkUj1plgbj59b4dC39QV2\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698990,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd396c0ab71698-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:56:30 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '951'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_2c3cb5caed61ccd1e058ef3e6301c691
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: !!binary |
+      Cq0TCiQKIgoMc2VydmljZS5uYW1lEhIKEGNyZXdBSS10ZWxlbWV0cnkShBMKEgoQY3Jld2FpLnRl
+      bGVtZXRyeRKkBwoQzBQBWCz+GLuI1awj3OPWrRIIGpT16t5bk6MqDENyZXcgQ3JlYXRlZDABOUBz
+      OyuEGSEYQYDBSCuEGSEYShsKDmNyZXdhaV92ZXJzaW9uEgkKBzAuMTAwLjBKGgoOcHl0aG9uX3Zl
+      cnNpb24SCAoGMy4xMi44Si4KCGNyZXdfa2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhj
+      NzQ2MjhjSjEKB2NyZXdfaWQSJgokMDE3NjQ5ZWMtYTBlMS00MzYxLWFlNjgtYzA1N2E3ZGM5YzI5
+      ShwKDGNyZXdfcHJvY2VzcxIMCgpzZXF1ZW50aWFsShEKC2NyZXdfbWVtb3J5EgIQAEoaChRjcmV3
+      X251bWJlcl9vZl90YXNrcxICGAFKGwoVY3Jld19udW1iZXJfb2ZfYWdlbnRzEgIYAUrRAgoLY3Jl
+      d19hZ2VudHMSwQIKvgJbeyJrZXkiOiAiYWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIyNDljNGM2NGEi
+      LCAiaWQiOiAiOGU3NzgyN2QtN2Y2OC00ZDA2LWI2YTctOWI4YjRkMGE0YzMzIiwgInJvbGUiOiAi
+      YmFzZV9hZ2VudCIsICJ2ZXJib3NlPyI6IGZhbHNlLCAibWF4X2l0ZXIiOiAyMCwgIm1heF9ycG0i
+      OiBudWxsLCAiZnVuY3Rpb25fY2FsbGluZ19sbG0iOiAiIiwgImxsbSI6ICJncHQtNG8tbWluaSIs
+      ICJkZWxlZ2F0aW9uX2VuYWJsZWQ/IjogZmFsc2UsICJhbGxvd19jb2RlX2V4ZWN1dGlvbj8iOiBm
+      YWxzZSwgIm1heF9yZXRyeV9saW1pdCI6IDIsICJ0b29sc19uYW1lcyI6IFtdfV1K/wEKCmNyZXdf
+      dGFza3MS8AEK7QFbeyJrZXkiOiAiMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDIiLCAi
+      aWQiOiAiOTJiZDIzMWYtYzAxMC00ZDI3LWIxNGYtZjE5NjEyZTBmZTkzIiwgImFzeW5jX2V4ZWN1
+      dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lucHV0PyI6IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJiYXNl
+      X2FnZW50IiwgImFnZW50X2tleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIs
+      ICJ0b29sc19uYW1lcyI6IFtdfV16AhgBhQEAAQAAEo4CChC22Au0eMkAAjV6cfU1NrNIEggxb1Bq
+      Xnll/ioMVGFzayBDcmVhdGVkMAE5IOJaK4QZIRhBwG5bK4QZIRhKLgoIY3Jld19rZXkSIgogZTU4
+      MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiQwMTc2NDllYy1hMGUx
+      LTQzNjEtYWU2OC1jMDU3YTdkYzljMjlKLgoIdGFza19rZXkSIgogMWIxNWVmMjM5MTViMjc1NWU4
+      OWEwZWMzYjI2YTEzZDJKMQoHdGFza19pZBImCiQ5MmJkMjMxZi1jMDEwLTRkMjctYjE0Zi1mMTk2
+      MTJlMGZlOTN6AhgBhQEAAQAAEqQHChC63jCLGR8RP8RmYiHrdNVeEggZ39ffmGm5xyoMQ3JldyBD
+      cmVhdGVkMAE5GFEe04QZIRhBELEq04QZIRhKGwoOY3Jld2FpX3ZlcnNpb24SCQoHMC4xMDAuMEoa
+      Cg5weXRob25fdmVyc2lvbhIICgYzLjEyLjhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVh
+      ZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiQ5MTY4YmQxNC0yN2Q2LTQ3NWMtODljOC01
+      NjJjOTAyMGIxOTBKHAoMY3Jld19wcm9jZXNzEgwKCnNlcXVlbnRpYWxKEQoLY3Jld19tZW1vcnkS
+      AhAAShoKFGNyZXdfbnVtYmVyX29mX3Rhc2tzEgIYAUobChVjcmV3X251bWJlcl9vZl9hZ2VudHMS
+      AhgBStECCgtjcmV3X2FnZW50cxLBAgq+Alt7ImtleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2
+      YjI0OWM0YzY0YSIsICJpZCI6ICI4ZTc3ODI3ZC03ZjY4LTRkMDYtYjZhNy05YjhiNGQwYTRjMzMi
+      LCAicm9sZSI6ICJiYXNlX2FnZW50IiwgInZlcmJvc2U/IjogZmFsc2UsICJtYXhfaXRlciI6IDIw
+      LCAibWF4X3JwbSI6IG51bGwsICJmdW5jdGlvbl9jYWxsaW5nX2xsbSI6ICIiLCAibGxtIjogImdw
+      dC00by1taW5pIiwgImRlbGVnYXRpb25fZW5hYmxlZD8iOiBmYWxzZSwgImFsbG93X2NvZGVfZXhl
+      Y3V0aW9uPyI6IGZhbHNlLCAibWF4X3JldHJ5X2xpbWl0IjogMiwgInRvb2xzX25hbWVzIjogW119
+      XUr/AQoKY3Jld190YXNrcxLwAQrtAVt7ImtleSI6ICIxYjE1ZWYyMzkxNWIyNzU1ZTg5YTBlYzNi
+      MjZhMTNkMiIsICJpZCI6ICI5MmJkMjMxZi1jMDEwLTRkMjctYjE0Zi1mMTk2MTJlMGZlOTMiLCAi
+      YXN5bmNfZXhlY3V0aW9uPyI6IGZhbHNlLCAiaHVtYW5faW5wdXQ/IjogZmFsc2UsICJhZ2VudF9y
+      b2xlIjogImJhc2VfYWdlbnQiLCAiYWdlbnRfa2V5IjogImFkMTUzMTYxYzVjNWE4NTZhYTBkMDZi
+      MjQ5YzRjNjRhIiwgInRvb2xzX25hbWVzIjogW119XXoCGAGFAQABAAASjgIKEOo6FGs7r9hHrN+f
+      qhMTUysSCJgbYV+vQMbCKgxUYXNrIENyZWF0ZWQwATlAxjrThBkhGEEYIDvThBkhGEouCghjcmV3
+      X2tleRIiCiBlNTgwNzAxZDUyZWI2NWFmZjI0ZWVmZTc4Yzc0NjI4Y0oxCgdjcmV3X2lkEiYKJDkx
+      NjhiZDE0LTI3ZDYtNDc1Yy04OWM4LTU2MmM5MDIwYjE5MEouCgh0YXNrX2tleRIiCiAxYjE1ZWYy
+      MzkxNWIyNzU1ZTg5YTBlYzNiMjZhMTNkMkoxCgd0YXNrX2lkEiYKJDkyYmQyMzFmLWMwMTAtNGQy
+      Ny1iMTRmLWYxOTYxMmUwZmU5M3oCGAGFAQABAAA=
+    headers:
+      Accept:
+      - '*/*'
+      Accept-Encoding:
+      - gzip, deflate
+      Connection:
+      - keep-alive
+      Content-Length:
+      - '2480'
+      Content-Type:
+      - application/x-protobuf
+      User-Agent:
+      - OTel-OTLP-Exporter-Python/1.27.0
+    method: POST
+    uri: https://telemetry.crewai.com:4319/v1/traces
+  response:
+    body:
+      string: "\n\0"
+    headers:
+      Content-Length:
+      - '2'
+      Content-Type:
+      - application/x-protobuf
+      Date:
+      - Tue, 04 Feb 2025 19:56:31 GMT
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=4s6sWmJ49B9F_wNc1STtdZF1nikfl6uN9_ov3Xzfa8U-1738698987-1.0.1.1-lmbRRS1MHrDbnU93Gh16CP3qNczxxIrQnyBU7vpHSwNf6PdmuWOHKd1mkl5SBx6rg7p1NLaNUMyqDDcE0Mvjzw;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJK3bJiyqGhPeqdCcCjoeNavGHrR\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698991,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_uAFkclWHIRqgrXFrQFcEoUIS\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Include additional
+      context for the greeting to make it more meaningful.\\\",\\\"Specify if you
+      want a casual or formal tone for greetings.\\\",\\\"Provide examples of variations
+      of the greeting if necessary.\\\"],\\\"quality\\\":10,\\\"entities\\\":[],\\\"relationships\\\":[]}\"\n
+      \           }\n          }\n        ],\n        \"refusal\": null\n      },\n
+      \     \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n  ],\n
+      \ \"usage\": {\n    \"prompt_tokens\": 273,\n    \"completion_tokens\": 50,\n
+      \   \"total_tokens\": 323,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\":
+      0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\": {\n
+      \     \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_bd83329f63\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd3973589f1698-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:56:32 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1408'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_519fd27ca3d5da4d541c4331654e0520
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_end_task_event.yaml

@@ -0,0 +1,357 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=nedOdWE1YnKQYt1kSbrcA.zhwa3bZDzmZqTOjZYER0c-1738700521-1.0.1.1-xQk9iXOvqvyXNhkIOgc8Ws2WYcT1mJFkDCvCC8xA5joFD8QfNrBIAr_Qs6sIxt2EzXyeFwBA6gA8ZgWApCHx0Q;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJlK2np8dMxYgsDIuyz2TSKKELWh\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738700682,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd62c1fdb0fa6a-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 20:24:42 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '326'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_22be86be6fd9d69ca8d310ef534e7bec
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: !!binary |
+      Cp0mCiQKIgoMc2VydmljZS5uYW1lEhIKEGNyZXdBSS10ZWxlbWV0cnkS9CUKEgoQY3Jld2FpLnRl
+      bGVtZXRyeRKkBwoQfRmsiIy66Scw43CcUw5ZiBIIfYDEhWJTOTkqDENyZXcgQ3JlYXRlZDABOcD6
+      XvsOGyEYQWC2bPsOGyEYShsKDmNyZXdhaV92ZXJzaW9uEgkKBzAuMTAwLjBKGgoOcHl0aG9uX3Zl
+      cnNpb24SCAoGMy4xMi44Si4KCGNyZXdfa2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhj
+      NzQ2MjhjSjEKB2NyZXdfaWQSJgokMDMzNTZiYmEtMzJmZC00OThmLTgxYTItYzc1ZDBkMzc2N2Qx
+      ShwKDGNyZXdfcHJvY2VzcxIMCgpzZXF1ZW50aWFsShEKC2NyZXdfbWVtb3J5EgIQAEoaChRjcmV3
+      X251bWJlcl9vZl90YXNrcxICGAFKGwoVY3Jld19udW1iZXJfb2ZfYWdlbnRzEgIYAUrRAgoLY3Jl
+      d19hZ2VudHMSwQIKvgJbeyJrZXkiOiAiYWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIyNDljNGM2NGEi
+      LCAiaWQiOiAiZjY5YTRmYTMtMzQ4OC00MmFhLTlhMTQtMGEzZmEyOWJmYjZjIiwgInJvbGUiOiAi
+      YmFzZV9hZ2VudCIsICJ2ZXJib3NlPyI6IGZhbHNlLCAibWF4X2l0ZXIiOiAyMCwgIm1heF9ycG0i
+      OiBudWxsLCAiZnVuY3Rpb25fY2FsbGluZ19sbG0iOiAiIiwgImxsbSI6ICJncHQtNG8tbWluaSIs
+      ICJkZWxlZ2F0aW9uX2VuYWJsZWQ/IjogZmFsc2UsICJhbGxvd19jb2RlX2V4ZWN1dGlvbj8iOiBm
+      YWxzZSwgIm1heF9yZXRyeV9saW1pdCI6IDIsICJ0b29sc19uYW1lcyI6IFtdfV1K/wEKCmNyZXdf
+      dGFza3MS8AEK7QFbeyJrZXkiOiAiMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDIiLCAi
+      aWQiOiAiNDllNzk0MmMtZTJiMy00YmE1LTg5MTUtMTYwYjQxMDU2ZmVlIiwgImFzeW5jX2V4ZWN1
+      dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lucHV0PyI6IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJiYXNl
+      X2FnZW50IiwgImFnZW50X2tleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIs
+      ICJ0b29sc19uYW1lcyI6IFtdfV16AhgBhQEAAQAAEo4CChCLgjutPq/3tkNyPqVjfbZhEgjhB7lb
+      clxzdyoMVGFzayBDcmVhdGVkMAE5CC9++w4bIRhBeMN++w4bIRhKLgoIY3Jld19rZXkSIgogZTU4
+      MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiQwMzM1NmJiYS0zMmZk
+      LTQ5OGYtODFhMi1jNzVkMGQzNzY3ZDFKLgoIdGFza19rZXkSIgogMWIxNWVmMjM5MTViMjc1NWU4
+      OWEwZWMzYjI2YTEzZDJKMQoHdGFza19pZBImCiQ0OWU3OTQyYy1lMmIzLTRiYTUtODkxNS0xNjBi
+      NDEwNTZmZWV6AhgBhQEAAQAAEqQHChC7SpRSs6eG9XFmYuMQgghQEghpZMlScOy2DyoMQ3JldyBD
+      cmVhdGVkMAE50O2cAA8bIRhBOFmrAA8bIRhKGwoOY3Jld2FpX3ZlcnNpb24SCQoHMC4xMDAuMEoa
+      Cg5weXRob25fdmVyc2lvbhIICgYzLjEyLjhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVh
+      ZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiRlYTQ1MWY2OS00Zjk3LTQ4MjYtOWNlYi04
+      NTAzMDk2MTQ2MDlKHAoMY3Jld19wcm9jZXNzEgwKCnNlcXVlbnRpYWxKEQoLY3Jld19tZW1vcnkS
+      AhAAShoKFGNyZXdfbnVtYmVyX29mX3Rhc2tzEgIYAUobChVjcmV3X251bWJlcl9vZl9hZ2VudHMS
+      AhgBStECCgtjcmV3X2FnZW50cxLBAgq+Alt7ImtleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2
+      YjI0OWM0YzY0YSIsICJpZCI6ICJmNjlhNGZhMy0zNDg4LTQyYWEtOWExNC0wYTNmYTI5YmZiNmMi
+      LCAicm9sZSI6ICJiYXNlX2FnZW50IiwgInZlcmJvc2U/IjogZmFsc2UsICJtYXhfaXRlciI6IDIw
+      LCAibWF4X3JwbSI6IG51bGwsICJmdW5jdGlvbl9jYWxsaW5nX2xsbSI6ICIiLCAibGxtIjogImdw
+      dC00by1taW5pIiwgImRlbGVnYXRpb25fZW5hYmxlZD8iOiBmYWxzZSwgImFsbG93X2NvZGVfZXhl
+      Y3V0aW9uPyI6IGZhbHNlLCAibWF4X3JldHJ5X2xpbWl0IjogMiwgInRvb2xzX25hbWVzIjogW119
+      XUr/AQoKY3Jld190YXNrcxLwAQrtAVt7ImtleSI6ICIxYjE1ZWYyMzkxNWIyNzU1ZTg5YTBlYzNi
+      MjZhMTNkMiIsICJpZCI6ICI0OWU3OTQyYy1lMmIzLTRiYTUtODkxNS0xNjBiNDEwNTZmZWUiLCAi
+      YXN5bmNfZXhlY3V0aW9uPyI6IGZhbHNlLCAiaHVtYW5faW5wdXQ/IjogZmFsc2UsICJhZ2VudF9y
+      b2xlIjogImJhc2VfYWdlbnQiLCAiYWdlbnRfa2V5IjogImFkMTUzMTYxYzVjNWE4NTZhYTBkMDZi
+      MjQ5YzRjNjRhIiwgInRvb2xzX25hbWVzIjogW119XXoCGAGFAQABAAASjgIKEM1q5zTl6Q6WKHPQ
+      eIqEm7sSCCWi2wvaFqpfKgxUYXNrIENyZWF0ZWQwATmIObwADxshGEEAo7wADxshGEouCghjcmV3
+      X2tleRIiCiBlNTgwNzAxZDUyZWI2NWFmZjI0ZWVmZTc4Yzc0NjI4Y0oxCgdjcmV3X2lkEiYKJGVh
+      NDUxZjY5LTRmOTctNDgyNi05Y2ViLTg1MDMwOTYxNDYwOUouCgh0YXNrX2tleRIiCiAxYjE1ZWYy
+      MzkxNWIyNzU1ZTg5YTBlYzNiMjZhMTNkMkoxCgd0YXNrX2lkEiYKJDQ5ZTc5NDJjLWUyYjMtNGJh
+      NS04OTE1LTE2MGI0MTA1NmZlZXoCGAGFAQABAAASpAcKEEhIZzGdZRAdvfcluDR5qvESCFMGo60X
+      V/dYKgxDcmV3IENyZWF0ZWQwATnY8LgBDxshGEGQrsUBDxshGEobCg5jcmV3YWlfdmVyc2lvbhIJ
+      CgcwLjEwMC4wShoKDnB5dGhvbl92ZXJzaW9uEggKBjMuMTIuOEouCghjcmV3X2tleRIiCiBlNTgw
+      NzAxZDUyZWI2NWFmZjI0ZWVmZTc4Yzc0NjI4Y0oxCgdjcmV3X2lkEiYKJDA3MDJmOGU1LWRjMTYt
+      NDlhYi1hMWY1LThjOWQyY2IwMDYwYkocCgxjcmV3X3Byb2Nlc3MSDAoKc2VxdWVudGlhbEoRCgtj
+      cmV3X21lbW9yeRICEABKGgoUY3Jld19udW1iZXJfb2ZfdGFza3MSAhgBShsKFWNyZXdfbnVtYmVy
+      X29mX2FnZW50cxICGAFK0QIKC2NyZXdfYWdlbnRzEsECCr4CW3sia2V5IjogImFkMTUzMTYxYzVj
+      NWE4NTZhYTBkMDZiMjQ5YzRjNjRhIiwgImlkIjogImY2OWE0ZmEzLTM0ODgtNDJhYS05YTE0LTBh
+      M2ZhMjliZmI2YyIsICJyb2xlIjogImJhc2VfYWdlbnQiLCAidmVyYm9zZT8iOiBmYWxzZSwgIm1h
+      eF9pdGVyIjogMjAsICJtYXhfcnBtIjogbnVsbCwgImZ1bmN0aW9uX2NhbGxpbmdfbGxtIjogIiIs
+      ICJsbG0iOiAiZ3B0LTRvLW1pbmkiLCAiZGVsZWdhdGlvbl9lbmFibGVkPyI6IGZhbHNlLCAiYWxs
+      b3dfY29kZV9leGVjdXRpb24/IjogZmFsc2UsICJtYXhfcmV0cnlfbGltaXQiOiAyLCAidG9vbHNf
+      bmFtZXMiOiBbXX1dSv8BCgpjcmV3X3Rhc2tzEvABCu0BW3sia2V5IjogIjFiMTVlZjIzOTE1YjI3
+      NTVlODlhMGVjM2IyNmExM2QyIiwgImlkIjogIjQ5ZTc5NDJjLWUyYjMtNGJhNS04OTE1LTE2MGI0
+      MTA1NmZlZSIsICJhc3luY19leGVjdXRpb24/IjogZmFsc2UsICJodW1hbl9pbnB1dD8iOiBmYWxz
+      ZSwgImFnZW50X3JvbGUiOiAiYmFzZV9hZ2VudCIsICJhZ2VudF9rZXkiOiAiYWQxNTMxNjFjNWM1
+      YTg1NmFhMGQwNmIyNDljNGM2NGEiLCAidG9vbHNfbmFtZXMiOiBbXX1degIYAYUBAAEAABKOAgoQ
+      PkgvMtq3aT5YPXE8gCRxPBIIMx/zQOuC+8sqDFRhc2sgQ3JlYXRlZDABOYjk2AEPGyEYQZiI2QEP
+      GyEYSi4KCGNyZXdfa2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhjNzQ2MjhjSjEKB2Ny
+      ZXdfaWQSJgokMDcwMmY4ZTUtZGMxNi00OWFiLWExZjUtOGM5ZDJjYjAwNjBiSi4KCHRhc2tfa2V5
+      EiIKIDFiMTVlZjIzOTE1YjI3NTVlODlhMGVjM2IyNmExM2QySjEKB3Rhc2tfaWQSJgokNDllNzk0
+      MmMtZTJiMy00YmE1LTg5MTUtMTYwYjQxMDU2ZmVlegIYAYUBAAEAABKkBwoQex7nA0gUUrZHbN6F
+      gWp/gBIIKPd4fiRi7DwqDENyZXcgQ3JlYXRlZDABOTiIrAIPGyEYQdCAtwIPGyEYShsKDmNyZXdh
+      aV92ZXJzaW9uEgkKBzAuMTAwLjBKGgoOcHl0aG9uX3ZlcnNpb24SCAoGMy4xMi44Si4KCGNyZXdf
+      a2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhjNzQ2MjhjSjEKB2NyZXdfaWQSJgokMGNm
+      YjUzZWItMDA2Mi00YmVmLTk1ZTgtMDgwMjQ3NmNkMWRlShwKDGNyZXdfcHJvY2VzcxIMCgpzZXF1
+      ZW50aWFsShEKC2NyZXdfbWVtb3J5EgIQAEoaChRjcmV3X251bWJlcl9vZl90YXNrcxICGAFKGwoV
+      Y3Jld19udW1iZXJfb2ZfYWdlbnRzEgIYAUrRAgoLY3Jld19hZ2VudHMSwQIKvgJbeyJrZXkiOiAi
+      YWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIyNDljNGM2NGEiLCAiaWQiOiAiZjY5YTRmYTMtMzQ4OC00
+      MmFhLTlhMTQtMGEzZmEyOWJmYjZjIiwgInJvbGUiOiAiYmFzZV9hZ2VudCIsICJ2ZXJib3NlPyI6
+      IGZhbHNlLCAibWF4X2l0ZXIiOiAyMCwgIm1heF9ycG0iOiBudWxsLCAiZnVuY3Rpb25fY2FsbGlu
+      Z19sbG0iOiAiIiwgImxsbSI6ICJncHQtNG8tbWluaSIsICJkZWxlZ2F0aW9uX2VuYWJsZWQ/Ijog
+      ZmFsc2UsICJhbGxvd19jb2RlX2V4ZWN1dGlvbj8iOiBmYWxzZSwgIm1heF9yZXRyeV9saW1pdCI6
+      IDIsICJ0b29sc19uYW1lcyI6IFtdfV1K/wEKCmNyZXdfdGFza3MS8AEK7QFbeyJrZXkiOiAiMWIx
+      NWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDIiLCAiaWQiOiAiNDllNzk0MmMtZTJiMy00YmE1
+      LTg5MTUtMTYwYjQxMDU2ZmVlIiwgImFzeW5jX2V4ZWN1dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lu
+      cHV0PyI6IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJiYXNlX2FnZW50IiwgImFnZW50X2tleSI6ICJh
+      ZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIsICJ0b29sc19uYW1lcyI6IFtdfV16AhgB
+      hQEAAQAAEo4CChD+ymllejCksajGYDua8tgNEghnjlejrbuw2SoMVGFzayBDcmVhdGVkMAE5EDrI
+      Ag8bIRhBGIzIAg8bIRhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3
+      NDYyOGNKMQoHY3Jld19pZBImCiQwY2ZiNTNlYi0wMDYyLTRiZWYtOTVlOC0wODAyNDc2Y2QxZGVK
+      LgoIdGFza19rZXkSIgogMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDJKMQoHdGFza19p
+      ZBImCiQ0OWU3OTQyYy1lMmIzLTRiYTUtODkxNS0xNjBiNDEwNTZmZWV6AhgBhQEAAQAA
+    headers:
+      Accept:
+      - '*/*'
+      Accept-Encoding:
+      - gzip, deflate
+      Connection:
+      - keep-alive
+      Content-Length:
+      - '4896'
+      Content-Type:
+      - application/x-protobuf
+      User-Agent:
+      - OTel-OTLP-Exporter-Python/1.27.0
+    method: POST
+    uri: https://telemetry.crewai.com:4319/v1/traces
+  response:
+    body:
+      string: "\n\0"
+    headers:
+      Content-Length:
+      - '2'
+      Content-Type:
+      - application/x-protobuf
+      Date:
+      - Tue, 04 Feb 2025 20:24:47 GMT
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=nedOdWE1YnKQYt1kSbrcA.zhwa3bZDzmZqTOjZYER0c-1738700521-1.0.1.1-xQk9iXOvqvyXNhkIOgc8Ws2WYcT1mJFkDCvCC8xA5joFD8QfNrBIAr_Qs6sIxt2EzXyeFwBA6gA8ZgWApCHx0Q;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJlLVC3gCB9gRI0ZSkoPCZY7EwpQ\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738700683,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_mgwImOITW8lkjzAyf9Pp76cL\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Provide context or
+      additional information to make tasks more engaging.\\\",\\\"Encourage variations
+      in responses to make the interaction more dynamic.\\\"],\\\"quality\\\":10,\\\"entities\\\":[{\\\"name\\\":\\\"hi\\\",\\\"type\\\":\\\"greeting\\\",\\\"description\\\":\\\"A
+      common word used to initiate a conversation or express friendliness.\\\",\\\"relationships\\\":[\\\"initiates
+      conversation\\\",\\\"expresses friendliness\\\"]}]}\"\n            }\n          }\n
+      \       ],\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      273,\n    \"completion_tokens\": 71,\n    \"total_tokens\": 344,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_bd83329f63\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 90cd62c4ba41fa6a-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 20:24:50 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '7347'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_aec28dd3fe998d628754e8429623bf9e
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_kickoff_events.yaml

@@ -0,0 +1,245 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJIrSWAFqDEsNtLRhcM8vMHO9Ejw\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698917,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd37a83f5f176a-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:55:18 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=rKQWp4fbAvcCp4rasEN6DqiTjQfiWYpLfjcLpWcmzi0-1738698918-1.0.1.1-qlcCSdBY3KWbzVms0eLtz5ub5SSLGs_sRLxTdNhDk_purQuz9k6EFp8PHJfN3aP_sLnuyKnFlppM3.2k_dCtPQ;
+        path=/; expires=Tue, 04-Feb-25 20:25:18 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=Oi91zDXvjWohBYXSVqK4hFsq3_GZePEIIbi7b7wrjcI-1738698918130-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '894'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_864253996bbc0f797f9a2c1b9247a0d5
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=rKQWp4fbAvcCp4rasEN6DqiTjQfiWYpLfjcLpWcmzi0-1738698918-1.0.1.1-qlcCSdBY3KWbzVms0eLtz5ub5SSLGs_sRLxTdNhDk_purQuz9k6EFp8PHJfN3aP_sLnuyKnFlppM3.2k_dCtPQ;
+        _cfuvid=Oi91zDXvjWohBYXSVqK4hFsq3_GZePEIIbi7b7wrjcI-1738698918130-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJIsVEppA04iGQh0k6sanKnVObrO\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698918,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_AQ3iizjGWjEvk1SmhGCzjbf1\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Provide context for
+      the greeting, like a specific scenario or recipient.\\\",\\\"Encourage responses
+      or follow-ups to promote engagement.\\\",\\\"Specify the tone or formality of
+      the greeting, if relevant.\\\"],\\\"quality\\\":10,\\\"entities\\\":[{\\\"name\\\":\\\"hi\\\",\\\"type\\\":\\\"greeting\\\",\\\"description\\\":\\\"A
+      common informal expression used to initiate conversation or acknowledge someone.\\\",\\\"relationships\\\":[\\\"used
+      in conversation\\\",\\\"expresses friendliness\\\"]}]}\"\n            }\n          }\n
+      \       ],\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      273,\n    \"completion_tokens\": 84,\n    \"total_tokens\": 357,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_bd83329f63\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 90cd37aec8c8176a-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:55:21 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '3269'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_e6e67a3f5c6f2d48e0351cdce95edd97
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_start_kickoff_event.yaml

@@ -0,0 +1,243 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJJzafmayYpGTsTAWbOyZkmQJNa5\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698987,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 90cd395b0e641698-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:56:27 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=4s6sWmJ49B9F_wNc1STtdZF1nikfl6uN9_ov3Xzfa8U-1738698987-1.0.1.1-lmbRRS1MHrDbnU93Gh16CP3qNczxxIrQnyBU7vpHSwNf6PdmuWOHKd1mkl5SBx6rg7p1NLaNUMyqDDcE0Mvjzw;
+        path=/; expires=Tue, 04-Feb-25 20:26:27 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '839'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_577b484a927b455c40ed80f9fd4d9106
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=4s6sWmJ49B9F_wNc1STtdZF1nikfl6uN9_ov3Xzfa8U-1738698987-1.0.1.1-lmbRRS1MHrDbnU93Gh16CP3qNczxxIrQnyBU7vpHSwNf6PdmuWOHKd1mkl5SBx6rg7p1NLaNUMyqDDcE0Mvjzw;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJJz10KP7iadNPdKsbcsvHBa7cic\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738698987,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_czeHQgy5eiOVa0zlrtcfwepe\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Provide more context
+      or details for similar tasks to enhance output expectations.\\\",\\\"Encourage
+      creativity in responses for simple tasks to engage users more effectively.\\\"],\\\"quality\\\":10,\\\"entities\\\":[]
+      }\"\n            }\n          }\n        ],\n        \"refusal\": null\n      },\n
+      \     \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n  ],\n
+      \ \"usage\": {\n    \"prompt_tokens\": 273,\n    \"completion_tokens\": 40,\n
+      \   \"total_tokens\": 313,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\":
+      0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\": {\n
+      \     \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_bd83329f63\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd39615b281698-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 19:56:29 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1411'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_3e717a80c7d9c5ea19893dd990aaae26
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_start_task_event.yaml

@@ -0,0 +1,245 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=4s6sWmJ49B9F_wNc1STtdZF1nikfl6uN9_ov3Xzfa8U-1738698987-1.0.1.1-lmbRRS1MHrDbnU93Gh16CP3qNczxxIrQnyBU7vpHSwNf6PdmuWOHKd1mkl5SBx6rg7p1NLaNUMyqDDcE0Mvjzw;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJiiHEQwIXsiG0Sd5wofcuhxVbo9\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738700520,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd5ecd0f7667ee-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 20:22:01 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=nedOdWE1YnKQYt1kSbrcA.zhwa3bZDzmZqTOjZYER0c-1738700521-1.0.1.1-xQk9iXOvqvyXNhkIOgc8Ws2WYcT1mJFkDCvCC8xA5joFD8QfNrBIAr_Qs6sIxt2EzXyeFwBA6gA8ZgWApCHx0Q;
+        path=/; expires=Tue, 04-Feb-25 20:52:01 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '450'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_10eaafc81640a98a0a4789d270dd94d9
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1962'
+      content-type:
+      - application/json
+      cookie:
+      - __cf_bm=nedOdWE1YnKQYt1kSbrcA.zhwa3bZDzmZqTOjZYER0c-1738700521-1.0.1.1-xQk9iXOvqvyXNhkIOgc8Ws2WYcT1mJFkDCvCC8xA5joFD8QfNrBIAr_Qs6sIxt2EzXyeFwBA6gA8ZgWApCHx0Q;
+        _cfuvid=Cl48aI8.jSRja0Pqr6Jrh3mAnigd4rDn6lhGicyjMPY-1738698987673-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AxJijOhk12Ua6lS23IwtZTachfjq9\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1738700521,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_DSteeMHHPf5RanJb8qjCo4qx\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Consider adding context
+      for the greeting to make it more engaging.\\\",\\\"Specify if any additional
+      information or tone is desired in the greeting.\\\"],\\\"quality\\\":10,\\\"entities\\\":[{\\\"name\\\":\\\"greeting\\\",\\\"type\\\":\\\"text\\\",\\\"description\\\":\\\"A
+      simple greeting phrase\\\",\\\"relationships\\\":[\\\"is a\\\",\\\"is part of
+      a conversation\\\"]}]}\"\n            }\n          }\n        ],\n        \"refusal\":
+      null\n      },\n      \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n
+      \   }\n  ],\n  \"usage\": {\n    \"prompt_tokens\": 273,\n    \"completion_tokens\":
+      67,\n    \"total_tokens\": 340,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\":
+      0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\": {\n
+      \     \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_bd83329f63\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90cd5ed20cb267ee-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 04 Feb 2025 20:22:02 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1624'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999876'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_4ee944acdd3928afbf6c5562403b064a
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_crew_emits_task_failed_event.yaml

@@ -0,0 +1,114 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzpkZLpCyjKT5d6Udfx4zAme2sOMy\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739300299,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 910691d3ab90ebef-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 11 Feb 2025 18:58:20 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=MOH5EY6n3p8JKY53.yz7qzLuLYsEB8QdQXH09loUMBM-1739300300-1.0.1.1-hjb4mk04sMygPFhoFyiySKZSqB_fN5PbhbOyn.kipa3.eLvk7EtriDyjvGkBFIAV13DYnc08BfF_l2kxdx9hfQ;
+        path=/; expires=Tue, 11-Feb-25 19:28:20 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=uu.cEiV.FfgvSvCdKOooDYJWrwjVEuFeGdQodijGUUI-1739300300232-0.0.1.1-604800000;
+        path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1357'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_2277503f851195e7d7a43b66eb044454
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_multiple_handlers_for_same_event.yaml

@@ -0,0 +1,111 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=gsNyCo_jrDOolzf8SXHDaxQQrEgdR3jgv4OAH8MziDE-1739291824699-0.0.1.1-604800000;
+        __cf_bm=cRijYuylMGzRGxv3udQL5PhHOR5mRN_9_eLLwevlM_o-1739299455-1.0.1.1-Fszr_Msw0B1.IBMkiunP.VF2ilul1YGZZV8TqMcO3Q2SHvSlqfgm9NHgns1bJrm0wWRvHiCE7wdZfUAOx7T3Lg
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzpWx6pctOvzu6xsbyg0XfSAc0q9V\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739299455,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 91067d3ddc68fa16-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 11 Feb 2025 18:44:16 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '703'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_89222c00e4608e8557a135e91b223556
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_register_handler_adds_new_handler.yaml

@@ -0,0 +1,114 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nTo
+      give my best complete final answer to the task respond using the exact following
+      format:\n\nThought: I now can give a great answer\nFinal Answer: Your final
+      answer must be the great and the most complete as possible, it must be outcome
+      described.\n\nI MUST use these formats, my job depends on it!"}, {"role": "user",
+      "content": "\nCurrent Task: Just say hi\n\nThis is the expect criteria for your
+      final answer: hi\nyou MUST return the actual complete content as the final answer,
+      not a summary.\n\nBegin! This is VERY important to you, use the tools available
+      and give your best Final Answer, your job depends on it!\n\nThought:"}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '836'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=gsNyCo_jrDOolzf8SXHDaxQQrEgdR3jgv4OAH8MziDE-1739291824699-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzpWxLzAcRzigZuIGmjP3ckQgxAom\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739299455,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"I now can give a great answer  \\nFinal
+      Answer: hi\",\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      161,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 173,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 91067d389e90fa16-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Tue, 11 Feb 2025 18:44:15 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=cRijYuylMGzRGxv3udQL5PhHOR5mRN_9_eLLwevlM_o-1739299455-1.0.1.1-Fszr_Msw0B1.IBMkiunP.VF2ilul1YGZZV8TqMcO3Q2SHvSlqfgm9NHgns1bJrm0wWRvHiCE7wdZfUAOx7T3Lg;
+        path=/; expires=Tue, 11-Feb-25 19:14:15 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '716'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999810'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_ef807dc3223d40332aae8a313e96ef3a
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/cassettes/test_tools_emits_finished_events.yaml

@@ -0,0 +1,512 @@
+interactions:
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nYou
+      ONLY have access to the following tools, and should NEVER make up tools that
+      are not listed here:\n\nTool Name: say_hi\nTool Arguments: {}\nTool Description:
+      Say hi\n\nIMPORTANT: Use the following format in your response:\n\n```\nThought:
+      you should always think about what to do\nAction: the action to take, only one
+      name of [say_hi], just the name, exactly as it''s written.\nAction Input: the
+      input to the action, just a simple JSON object, enclosed in curly braces, using
+      \" to wrap keys and values.\nObservation: the result of the action\n```\n\nOnce
+      all necessary information is gathered, return the following format:\n\n```\nThought:
+      I now know the final answer\nFinal Answer: the final answer to the original
+      input question\n```"}, {"role": "user", "content": "\nCurrent Task: Just say
+      hi\n\nThis is the expect criteria for your final answer: hi\nyou MUST return
+      the actual complete content as the final answer, not a summary.\n\nBegin! This
+      is VERY important to you, use the tools available and give your best Final Answer,
+      your job depends on it!\n\nThought:"}], "model": "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1275'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=efIHP1NUsh1dFewGJBu4YoBu6hhGa8vjOOKQglYQGno-1739214901306-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzUA6kJQfpUvB4CGot4gSfAIR0foh\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739217314,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"you should always think about what to
+      do  \\nAction: say_hi  \\nAction Input: {}  \",\n        \"refusal\": null\n
+      \     },\n      \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n
+      \ ],\n  \"usage\": {\n    \"prompt_tokens\": 257,\n    \"completion_tokens\":
+      19,\n    \"total_tokens\": 276,\n    \"prompt_tokens_details\": {\n      \"cached_tokens\":
+      0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\": {\n
+      \     \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90fea7d78e1fceb9-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Mon, 10 Feb 2025 19:55:15 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=fmlg1wjOwuOwZhUUOEtL1tQYluAPumn7AHLF8s0EU2Y-1739217315-1.0.1.1-PQDvxn8TOhzaznlHjwVsqPZUzbAyJWFkvzCubfNJydTu2_AyA1cJ8hkM0khsEE4UY_xp8iPe2gSGmH1ydrDa0Q;
+        path=/; expires=Mon, 10-Feb-25 20:25:15 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '526'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999703'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_f6358ff0cc7a2b8d2e167ab00a40f2a4
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are base_agent. You are
+      a helpful assistant that just says hi\nYour personal goal is: Just say hi\nYou
+      ONLY have access to the following tools, and should NEVER make up tools that
+      are not listed here:\n\nTool Name: say_hi\nTool Arguments: {}\nTool Description:
+      Say hi\n\nIMPORTANT: Use the following format in your response:\n\n```\nThought:
+      you should always think about what to do\nAction: the action to take, only one
+      name of [say_hi], just the name, exactly as it''s written.\nAction Input: the
+      input to the action, just a simple JSON object, enclosed in curly braces, using
+      \" to wrap keys and values.\nObservation: the result of the action\n```\n\nOnce
+      all necessary information is gathered, return the following format:\n\n```\nThought:
+      I now know the final answer\nFinal Answer: the final answer to the original
+      input question\n```"}, {"role": "user", "content": "\nCurrent Task: Just say
+      hi\n\nThis is the expect criteria for your final answer: hi\nyou MUST return
+      the actual complete content as the final answer, not a summary.\n\nBegin! This
+      is VERY important to you, use the tools available and give your best Final Answer,
+      your job depends on it!\n\nThought:"}, {"role": "assistant", "content": "you
+      should always think about what to do  \nAction: say_hi  \nAction Input: {}  \nObservation:
+      hi"}], "model": "gpt-4o-mini", "stop": ["\nObservation:"]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1410'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=efIHP1NUsh1dFewGJBu4YoBu6hhGa8vjOOKQglYQGno-1739214901306-0.0.1.1-604800000;
+        __cf_bm=fmlg1wjOwuOwZhUUOEtL1tQYluAPumn7AHLF8s0EU2Y-1739217315-1.0.1.1-PQDvxn8TOhzaznlHjwVsqPZUzbAyJWFkvzCubfNJydTu2_AyA1cJ8hkM0khsEE4UY_xp8iPe2gSGmH1ydrDa0Q
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzUA7QdlQy1WZZijxNWUv25sZycg0\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739217315,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"```\\nThought: I now know the final answer\\nFinal
+      Answer: hi\\n```\",\n        \"refusal\": null\n      },\n      \"logprobs\":
+      null,\n      \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      283,\n    \"completion_tokens\": 17,\n    \"total_tokens\": 300,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90fea7dc5ba6ceb9-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Mon, 10 Feb 2025 19:55:15 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '388'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999680'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_7d7c68b90b3a9c3ac6092fe17ac1185a
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: !!binary |
+      CoMzCiQKIgoMc2VydmljZS5uYW1lEhIKEGNyZXdBSS10ZWxlbWV0cnkS2jIKEgoQY3Jld2FpLnRl
+      bGVtZXRyeRKOAgoQ2EINIGZRoXD589od63oHmBIIMfUgEWudUbIqDFRhc2sgQ3JlYXRlZDABOcjI
+      7lbu8CIYQZB471bu8CIYSi4KCGNyZXdfa2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhj
+      NzQ2MjhjSjEKB2NyZXdfaWQSJgokNTE4ODdiOTktY2FlMy00Yjc4LWJjMGEtMDY4MmVmNWEzNGQ0
+      Si4KCHRhc2tfa2V5EiIKIDFiMTVlZjIzOTE1YjI3NTVlODlhMGVjM2IyNmExM2QySjEKB3Rhc2tf
+      aWQSJgokMzlmMDlmMWUtOTJmOC00ZGJiLTgzNDAtNjU2ZmVkMDk3ZjM0egIYAYUBAAEAABKkBwoQ
+      RzhWoF6ewSTS/qUc9yeFRhIIM3SNZCwjz5AqDENyZXcgQ3JlYXRlZDABOQjrGlru8CIYQdgbKVru
+      8CIYShsKDmNyZXdhaV92ZXJzaW9uEgkKBzAuMTAwLjBKGgoOcHl0aG9uX3ZlcnNpb24SCAoGMy4x
+      Mi44Si4KCGNyZXdfa2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhjNzQ2MjhjSjEKB2Ny
+      ZXdfaWQSJgokYzk4ODFkY2YtMmM0MS00ZjRlLTgzMjctNjJjYjFhYjJkOTg4ShwKDGNyZXdfcHJv
+      Y2VzcxIMCgpzZXF1ZW50aWFsShEKC2NyZXdfbWVtb3J5EgIQAEoaChRjcmV3X251bWJlcl9vZl90
+      YXNrcxICGAFKGwoVY3Jld19udW1iZXJfb2ZfYWdlbnRzEgIYAUrRAgoLY3Jld19hZ2VudHMSwQIK
+      vgJbeyJrZXkiOiAiYWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIyNDljNGM2NGEiLCAiaWQiOiAiNTU2
+      NzJiMDgtOTU4ZC00MjljLWE3ZTctY2ZlN2U4Y2MwOGZkIiwgInJvbGUiOiAiYmFzZV9hZ2VudCIs
+      ICJ2ZXJib3NlPyI6IGZhbHNlLCAibWF4X2l0ZXIiOiAyMCwgIm1heF9ycG0iOiBudWxsLCAiZnVu
+      Y3Rpb25fY2FsbGluZ19sbG0iOiAiIiwgImxsbSI6ICJncHQtNG8tbWluaSIsICJkZWxlZ2F0aW9u
+      X2VuYWJsZWQ/IjogZmFsc2UsICJhbGxvd19jb2RlX2V4ZWN1dGlvbj8iOiBmYWxzZSwgIm1heF9y
+      ZXRyeV9saW1pdCI6IDIsICJ0b29sc19uYW1lcyI6IFtdfV1K/wEKCmNyZXdfdGFza3MS8AEK7QFb
+      eyJrZXkiOiAiMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDIiLCAiaWQiOiAiMzlmMDlm
+      MWUtOTJmOC00ZGJiLTgzNDAtNjU2ZmVkMDk3ZjM0IiwgImFzeW5jX2V4ZWN1dGlvbj8iOiBmYWxz
+      ZSwgImh1bWFuX2lucHV0PyI6IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJiYXNlX2FnZW50IiwgImFn
+      ZW50X2tleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIsICJ0b29sc19uYW1l
+      cyI6IFtdfV16AhgBhQEAAQAAEo4CChB8AxWkb2Uwpdc8RpyCRqw5EggJAxbgNu81XyoMVGFzayBD
+      cmVhdGVkMAE5+HQ8Wu7wIhhB+PE8Wu7wIhhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVh
+      ZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiRjOTg4MWRjZi0yYzQxLTRmNGUtODMyNy02
+      MmNiMWFiMmQ5ODhKLgoIdGFza19rZXkSIgogMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEz
+      ZDJKMQoHdGFza19pZBImCiQzOWYwOWYxZS05MmY4LTRkYmItODM0MC02NTZmZWQwOTdmMzR6AhgB
+      hQEAAQAAEqQHChCcXvdbsgYC+gzCMrXs3LN/EgijKwJLCRIiHioMQ3JldyBDcmVhdGVkMAE5iJqz
+      vu7wIhhBqKC/vu7wIhhKGwoOY3Jld2FpX3ZlcnNpb24SCQoHMC4xMDAuMEoaCg5weXRob25fdmVy
+      c2lvbhIICgYzLjEyLjhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3
+      NDYyOGNKMQoHY3Jld19pZBImCiQ2Zjk1ZWI3Yy0wOWM5LTQxOTYtYWFiYi1kOWIxNmMxMzZjODdK
+      HAoMY3Jld19wcm9jZXNzEgwKCnNlcXVlbnRpYWxKEQoLY3Jld19tZW1vcnkSAhAAShoKFGNyZXdf
+      bnVtYmVyX29mX3Rhc2tzEgIYAUobChVjcmV3X251bWJlcl9vZl9hZ2VudHMSAhgBStECCgtjcmV3
+      X2FnZW50cxLBAgq+Alt7ImtleSI6ICJhZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIs
+      ICJpZCI6ICI1NTY3MmIwOC05NThkLTQyOWMtYTdlNy1jZmU3ZThjYzA4ZmQiLCAicm9sZSI6ICJi
+      YXNlX2FnZW50IiwgInZlcmJvc2U/IjogZmFsc2UsICJtYXhfaXRlciI6IDIwLCAibWF4X3JwbSI6
+      IG51bGwsICJmdW5jdGlvbl9jYWxsaW5nX2xsbSI6ICIiLCAibGxtIjogImdwdC00by1taW5pIiwg
+      ImRlbGVnYXRpb25fZW5hYmxlZD8iOiBmYWxzZSwgImFsbG93X2NvZGVfZXhlY3V0aW9uPyI6IGZh
+      bHNlLCAibWF4X3JldHJ5X2xpbWl0IjogMiwgInRvb2xzX25hbWVzIjogW119XUr/AQoKY3Jld190
+      YXNrcxLwAQrtAVt7ImtleSI6ICIxYjE1ZWYyMzkxNWIyNzU1ZTg5YTBlYzNiMjZhMTNkMiIsICJp
+      ZCI6ICIzOWYwOWYxZS05MmY4LTRkYmItODM0MC02NTZmZWQwOTdmMzQiLCAiYXN5bmNfZXhlY3V0
+      aW9uPyI6IGZhbHNlLCAiaHVtYW5faW5wdXQ/IjogZmFsc2UsICJhZ2VudF9yb2xlIjogImJhc2Vf
+      YWdlbnQiLCAiYWdlbnRfa2V5IjogImFkMTUzMTYxYzVjNWE4NTZhYTBkMDZiMjQ5YzRjNjRhIiwg
+      InRvb2xzX25hbWVzIjogW119XXoCGAGFAQABAAASjgIKEExDo5nPLyHb2H8DfYjPoX4SCLEYs+24
+      8EenKgxUYXNrIENyZWF0ZWQwATmI4NG+7vAiGEFYZdK+7vAiGEouCghjcmV3X2tleRIiCiBlNTgw
+      NzAxZDUyZWI2NWFmZjI0ZWVmZTc4Yzc0NjI4Y0oxCgdjcmV3X2lkEiYKJDZmOTVlYjdjLTA5Yzkt
+      NDE5Ni1hYWJiLWQ5YjE2YzEzNmM4N0ouCgh0YXNrX2tleRIiCiAxYjE1ZWYyMzkxNWIyNzU1ZTg5
+      YTBlYzNiMjZhMTNkMkoxCgd0YXNrX2lkEiYKJDM5ZjA5ZjFlLTkyZjgtNGRiYi04MzQwLTY1NmZl
+      ZDA5N2YzNHoCGAGFAQABAAASpAcKEBBQzR2bcR/7woQ+VkaJ4kQSCD1LFx3SNPPPKgxDcmV3IENy
+      ZWF0ZWQwATlotsW/7vAiGEEgA9C/7vAiGEobCg5jcmV3YWlfdmVyc2lvbhIJCgcwLjEwMC4wShoK
+      DnB5dGhvbl92ZXJzaW9uEggKBjMuMTIuOEouCghjcmV3X2tleRIiCiBlNTgwNzAxZDUyZWI2NWFm
+      ZjI0ZWVmZTc4Yzc0NjI4Y0oxCgdjcmV3X2lkEiYKJDJiMWI2MGYzLTNlZTMtNGNjYi05MDM2LTdk
+      MzE4OTJiYjVkZkocCgxjcmV3X3Byb2Nlc3MSDAoKc2VxdWVudGlhbEoRCgtjcmV3X21lbW9yeRIC
+      EABKGgoUY3Jld19udW1iZXJfb2ZfdGFza3MSAhgBShsKFWNyZXdfbnVtYmVyX29mX2FnZW50cxIC
+      GAFK0QIKC2NyZXdfYWdlbnRzEsECCr4CW3sia2V5IjogImFkMTUzMTYxYzVjNWE4NTZhYTBkMDZi
+      MjQ5YzRjNjRhIiwgImlkIjogIjU1NjcyYjA4LTk1OGQtNDI5Yy1hN2U3LWNmZTdlOGNjMDhmZCIs
+      ICJyb2xlIjogImJhc2VfYWdlbnQiLCAidmVyYm9zZT8iOiBmYWxzZSwgIm1heF9pdGVyIjogMjAs
+      ICJtYXhfcnBtIjogbnVsbCwgImZ1bmN0aW9uX2NhbGxpbmdfbGxtIjogIiIsICJsbG0iOiAiZ3B0
+      LTRvLW1pbmkiLCAiZGVsZWdhdGlvbl9lbmFibGVkPyI6IGZhbHNlLCAiYWxsb3dfY29kZV9leGVj
+      dXRpb24/IjogZmFsc2UsICJtYXhfcmV0cnlfbGltaXQiOiAyLCAidG9vbHNfbmFtZXMiOiBbXX1d
+      Sv8BCgpjcmV3X3Rhc2tzEvABCu0BW3sia2V5IjogIjFiMTVlZjIzOTE1YjI3NTVlODlhMGVjM2Iy
+      NmExM2QyIiwgImlkIjogIjM5ZjA5ZjFlLTkyZjgtNGRiYi04MzQwLTY1NmZlZDA5N2YzNCIsICJh
+      c3luY19leGVjdXRpb24/IjogZmFsc2UsICJodW1hbl9pbnB1dD8iOiBmYWxzZSwgImFnZW50X3Jv
+      bGUiOiAiYmFzZV9hZ2VudCIsICJhZ2VudF9rZXkiOiAiYWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIy
+      NDljNGM2NGEiLCAidG9vbHNfbmFtZXMiOiBbXX1degIYAYUBAAEAABKOAgoQmT07KMiFRgzOOPQf
+      I4bJPhIIqzN+pCYM6IUqDFRhc2sgQ3JlYXRlZDABOYjr3r/u8CIYQehY37/u8CIYSi4KCGNyZXdf
+      a2V5EiIKIGU1ODA3MDFkNTJlYjY1YWZmMjRlZWZlNzhjNzQ2MjhjSjEKB2NyZXdfaWQSJgokMmIx
+      YjYwZjMtM2VlMy00Y2NiLTkwMzYtN2QzMTg5MmJiNWRmSi4KCHRhc2tfa2V5EiIKIDFiMTVlZjIz
+      OTE1YjI3NTVlODlhMGVjM2IyNmExM2QySjEKB3Rhc2tfaWQSJgokMzlmMDlmMWUtOTJmOC00ZGJi
+      LTgzNDAtNjU2ZmVkMDk3ZjM0egIYAYUBAAEAABKkBwoQE53vZNAWshkoNK1bqTvovRII83djkBUL
+      EbcqDENyZXcgQ3JlYXRlZDABORBBzsDu8CIYQbAU2MDu8CIYShsKDmNyZXdhaV92ZXJzaW9uEgkK
+      BzAuMTAwLjBKGgoOcHl0aG9uX3ZlcnNpb24SCAoGMy4xMi44Si4KCGNyZXdfa2V5EiIKIGU1ODA3
+      MDFkNTJlYjY1YWZmMjRlZWZlNzhjNzQ2MjhjSjEKB2NyZXdfaWQSJgokNTQ0MWY0MWYtOTVjMC00
+      YzdkLTkxM2QtNDUxODcwY2YyZjYzShwKDGNyZXdfcHJvY2VzcxIMCgpzZXF1ZW50aWFsShEKC2Ny
+      ZXdfbWVtb3J5EgIQAEoaChRjcmV3X251bWJlcl9vZl90YXNrcxICGAFKGwoVY3Jld19udW1iZXJf
+      b2ZfYWdlbnRzEgIYAUrRAgoLY3Jld19hZ2VudHMSwQIKvgJbeyJrZXkiOiAiYWQxNTMxNjFjNWM1
+      YTg1NmFhMGQwNmIyNDljNGM2NGEiLCAiaWQiOiAiNTU2NzJiMDgtOTU4ZC00MjljLWE3ZTctY2Zl
+      N2U4Y2MwOGZkIiwgInJvbGUiOiAiYmFzZV9hZ2VudCIsICJ2ZXJib3NlPyI6IGZhbHNlLCAibWF4
+      X2l0ZXIiOiAyMCwgIm1heF9ycG0iOiBudWxsLCAiZnVuY3Rpb25fY2FsbGluZ19sbG0iOiAiIiwg
+      ImxsbSI6ICJncHQtNG8tbWluaSIsICJkZWxlZ2F0aW9uX2VuYWJsZWQ/IjogZmFsc2UsICJhbGxv
+      d19jb2RlX2V4ZWN1dGlvbj8iOiBmYWxzZSwgIm1heF9yZXRyeV9saW1pdCI6IDIsICJ0b29sc19u
+      YW1lcyI6IFtdfV1K/wEKCmNyZXdfdGFza3MS8AEK7QFbeyJrZXkiOiAiMWIxNWVmMjM5MTViMjc1
+      NWU4OWEwZWMzYjI2YTEzZDIiLCAiaWQiOiAiMzlmMDlmMWUtOTJmOC00ZGJiLTgzNDAtNjU2ZmVk
+      MDk3ZjM0IiwgImFzeW5jX2V4ZWN1dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lucHV0PyI6IGZhbHNl
+      LCAiYWdlbnRfcm9sZSI6ICJiYXNlX2FnZW50IiwgImFnZW50X2tleSI6ICJhZDE1MzE2MWM1YzVh
+      ODU2YWEwZDA2YjI0OWM0YzY0YSIsICJ0b29sc19uYW1lcyI6IFtdfV16AhgBhQEAAQAAEo4CChBV
+      JNEz3VIdOlQM9VT3bctVEgisogN707a2AioMVGFzayBDcmVhdGVkMAE5kGbnwO7wIhhBaMDnwO7w
+      IhhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jl
+      d19pZBImCiQ1NDQxZjQxZi05NWMwLTRjN2QtOTEzZC00NTE4NzBjZjJmNjNKLgoIdGFza19rZXkS
+      IgogMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2YTEzZDJKMQoHdGFza19pZBImCiQzOWYwOWYx
+      ZS05MmY4LTRkYmItODM0MC02NTZmZWQwOTdmMzR6AhgBhQEAAQAAErQHChDA7zaLCfy56rd5t3oS
+      rDPZEgjYoSW3mq6WJyoMQ3JldyBDcmVhdGVkMAE5cP/5we7wIhhBIH0Dwu7wIhhKGwoOY3Jld2Fp
+      X3ZlcnNpb24SCQoHMC4xMDAuMEoaCg5weXRob25fdmVyc2lvbhIICgYzLjEyLjhKLgoIY3Jld19r
+      ZXkSIgogZTU4MDcwMWQ1MmViNjVhZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiRmNjcz
+      MTc1ZS04Y2Q1LTQ1ZWUtYTZiOS0xYWFjMTliODQxZWJKHAoMY3Jld19wcm9jZXNzEgwKCnNlcXVl
+      bnRpYWxKEQoLY3Jld19tZW1vcnkSAhAAShoKFGNyZXdfbnVtYmVyX29mX3Rhc2tzEgIYAUobChVj
+      cmV3X251bWJlcl9vZl9hZ2VudHMSAhgBStkCCgtjcmV3X2FnZW50cxLJAgrGAlt7ImtleSI6ICJh
+      ZDE1MzE2MWM1YzVhODU2YWEwZDA2YjI0OWM0YzY0YSIsICJpZCI6ICJmMGUwMGIzZi0wZWNmLTQ2
+      OGQtYjdjMC0yZmJhN2I5OTc5YjMiLCAicm9sZSI6ICJiYXNlX2FnZW50IiwgInZlcmJvc2U/Ijog
+      ZmFsc2UsICJtYXhfaXRlciI6IDIwLCAibWF4X3JwbSI6IG51bGwsICJmdW5jdGlvbl9jYWxsaW5n
+      X2xsbSI6ICIiLCAibGxtIjogImdwdC00by1taW5pIiwgImRlbGVnYXRpb25fZW5hYmxlZD8iOiBm
+      YWxzZSwgImFsbG93X2NvZGVfZXhlY3V0aW9uPyI6IGZhbHNlLCAibWF4X3JldHJ5X2xpbWl0Ijog
+      MiwgInRvb2xzX25hbWVzIjogWyJzYXlfaGkiXX1dSocCCgpjcmV3X3Rhc2tzEvgBCvUBW3sia2V5
+      IjogIjFiMTVlZjIzOTE1YjI3NTVlODlhMGVjM2IyNmExM2QyIiwgImlkIjogImFhMGFmMmE2LTdm
+      MTktNDZmNi1iMjMxLTg1M2JjYzYxYzhiZiIsICJhc3luY19leGVjdXRpb24/IjogZmFsc2UsICJo
+      dW1hbl9pbnB1dD8iOiBmYWxzZSwgImFnZW50X3JvbGUiOiAiYmFzZV9hZ2VudCIsICJhZ2VudF9r
+      ZXkiOiAiYWQxNTMxNjFjNWM1YTg1NmFhMGQwNmIyNDljNGM2NGEiLCAidG9vbHNfbmFtZXMiOiBb
+      InNheV9oaSJdfV16AhgBhQEAAQAAEo4CChBH8NUZY1Cv8sM2lfQLaEogEgiFlW7Wp7QpdyoMVGFz
+      ayBDcmVhdGVkMAE5MNkPwu7wIhhBUCcQwu7wIhhKLgoIY3Jld19rZXkSIgogZTU4MDcwMWQ1MmVi
+      NjVhZmYyNGVlZmU3OGM3NDYyOGNKMQoHY3Jld19pZBImCiRmNjczMTc1ZS04Y2Q1LTQ1ZWUtYTZi
+      OS0xYWFjMTliODQxZWJKLgoIdGFza19rZXkSIgogMWIxNWVmMjM5MTViMjc1NWU4OWEwZWMzYjI2
+      YTEzZDJKMQoHdGFza19pZBImCiRhYTBhZjJhNi03ZjE5LTQ2ZjYtYjIzMS04NTNiY2M2MWM4YmZ6
+      AhgBhQEAAQAAEooBChCJg/wSACw+HIDy4vvYISP/EgjoC/oI/1V0cCoKVG9vbCBVc2FnZTABOWA0
+      ifTu8CIYQTD0lPTu8CIYShsKDmNyZXdhaV92ZXJzaW9uEgkKBzAuMTAwLjBKFQoJdG9vbF9uYW1l
+      EggKBnNheV9oaUoOCghhdHRlbXB0cxICGAF6AhgBhQEAAQAA
+    headers:
+      Accept:
+      - '*/*'
+      Accept-Encoding:
+      - gzip, deflate
+      Connection:
+      - keep-alive
+      Content-Length:
+      - '6534'
+      Content-Type:
+      - application/x-protobuf
+      User-Agent:
+      - OTel-OTLP-Exporter-Python/1.27.0
+    method: POST
+    uri: https://telemetry.crewai.com:4319/v1/traces
+  response:
+    body:
+      string: "\n\0"
+    headers:
+      Content-Length:
+      - '2'
+      Content-Type:
+      - application/x-protobuf
+      Date:
+      - Mon, 10 Feb 2025 19:55:17 GMT
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "user", "content": "Assess the quality of the task
+      completed based on the description, expected output, and actual results.\n\nTask
+      Description:\nJust say hi\n\nExpected Output:\nhi\n\nActual Output:\nhi\n```\n\nPlease
+      provide:\n- Bullet points suggestions to improve future similar tasks\n- A score
+      from 0 to 10 evaluating on completion, quality, and overall performance- Entities
+      extracted from the task output, if any, their type, description, and relationships"}],
+      "model": "gpt-4o-mini", "tool_choice": {"type": "function", "function": {"name":
+      "TaskEvaluation"}}, "tools": [{"type": "function", "function": {"name": "TaskEvaluation",
+      "description": "Correctly extracted `TaskEvaluation` with all the required parameters
+      with correct types", "parameters": {"$defs": {"Entity": {"properties": {"name":
+      {"description": "The name of the entity.", "title": "Name", "type": "string"},
+      "type": {"description": "The type of the entity.", "title": "Type", "type":
+      "string"}, "description": {"description": "Description of the entity.", "title":
+      "Description", "type": "string"}, "relationships": {"description": "Relationships
+      of the entity.", "items": {"type": "string"}, "title": "Relationships", "type":
+      "array"}}, "required": ["name", "type", "description", "relationships"], "title":
+      "Entity", "type": "object"}}, "properties": {"suggestions": {"description":
+      "Suggestions to improve future similar tasks.", "items": {"type": "string"},
+      "title": "Suggestions", "type": "array"}, "quality": {"description": "A score
+      from 0 to 10 evaluating on completion, quality, and overall performance, all
+      taking into account the task description, expected output, and the result of
+      the task.", "title": "Quality", "type": "number"}, "entities": {"description":
+      "Entities extracted from the task output.", "items": {"$ref": "#/$defs/Entity"},
+      "title": "Entities", "type": "array"}}, "required": ["entities", "quality",
+      "suggestions"], "type": "object"}}}]}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '1967'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=efIHP1NUsh1dFewGJBu4YoBu6hhGa8vjOOKQglYQGno-1739214901306-0.0.1.1-604800000;
+        __cf_bm=fmlg1wjOwuOwZhUUOEtL1tQYluAPumn7AHLF8s0EU2Y-1739217315-1.0.1.1-PQDvxn8TOhzaznlHjwVsqPZUzbAyJWFkvzCubfNJydTu2_AyA1cJ8hkM0khsEE4UY_xp8iPe2gSGmH1ydrDa0Q
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.61.0
+      x-stainless-arch:
+      - arm64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - MacOS
+      x-stainless-package-version:
+      - 1.61.0
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.8
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AzUA8oE0A2d99i1Khpu0CI7fSgRtZ\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1739217316,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": null,\n        \"tool_calls\": [\n          {\n
+      \           \"id\": \"call_bk3duHRErK1qCyvWJ1uVmmGl\",\n            \"type\":
+      \"function\",\n            \"function\": {\n              \"name\": \"TaskEvaluation\",\n
+      \             \"arguments\": \"{\\\"suggestions\\\":[\\\"Provide more context
+      or details for similar tasks to enhance clarity.\\\",\\\"Specify desired tone
+      or style for the output.\\\",\\\"Consider adding more variety in tasks to keep
+      engagement high.\\\"],\\\"quality\\\":10,\\\"entities\\\":[{\\\"name\\\":\\\"hi\\\",\\\"type\\\":\\\"greeting\\\",\\\"description\\\":\\\"A
+      casual way to say hello or acknowledge someone's presence.\\\",\\\"relationships\\\":[\\\"used
+      as a greeting\\\",\\\"expresses friendliness\\\"]}]}\"\n            }\n          }\n
+      \       ],\n        \"refusal\": null\n      },\n      \"logprobs\": null,\n
+      \     \"finish_reason\": \"stop\"\n    }\n  ],\n  \"usage\": {\n    \"prompt_tokens\":
+      275,\n    \"completion_tokens\": 80,\n    \"total_tokens\": 355,\n    \"prompt_tokens_details\":
+      {\n      \"cached_tokens\": 0,\n      \"audio_tokens\": 0\n    },\n    \"completion_tokens_details\":
+      {\n      \"reasoning_tokens\": 0,\n      \"audio_tokens\": 0,\n      \"accepted_prediction_tokens\":
+      0,\n      \"rejected_prediction_tokens\": 0\n    }\n  },\n  \"service_tier\":
+      \"default\",\n  \"system_fingerprint\": \"fp_72ed7ab54c\"\n}\n"
+    headers:
+      CF-RAY:
+      - 90fea7dfef41ceb9-SJC
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Mon, 10 Feb 2025 19:55:17 GMT
+      Server:
+      - cloudflare
+      Transfer-Encoding:
+      - chunked
+      X-Content-Type-Options:
+      - nosniff
+      access-control-expose-headers:
+      - X-Request-ID
+      alt-svc:
+      - h3=":443"; ma=86400
+      cf-cache-status:
+      - DYNAMIC
+      openai-organization:
+      - crewai-iuxna1
+      openai-processing-ms:
+      - '1535'
+      openai-version:
+      - '2020-10-01'
+      strict-transport-security:
+      - max-age=31536000; includeSubDomains; preload
+      x-ratelimit-limit-requests:
+      - '30000'
+      x-ratelimit-limit-tokens:
+      - '150000000'
+      x-ratelimit-remaining-requests:
+      - '29999'
+      x-ratelimit-remaining-tokens:
+      - '149999874'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_55d8eb91b4318245556b73d3f4c1e7c4
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/utilities/test_datetime_compat.py

@@ -0,0 +1,35 @@
+"""Test datetime compatibility module."""
+from datetime import datetime, timedelta, timezone
+
+from crewai.utilities.datetime_compat import UTC
+
+
+def test_utc_timezone_compatibility():
+    """Test that UTC timezone is compatible with both Python 3.10 and 3.11+"""
+    assert UTC == timezone.utc
+    assert UTC.tzname(None) == "UTC"
+    # Verify it works with datetime.now()
+    dt = datetime.now(UTC)
+    assert dt.tzinfo == timezone.utc
+
+
+def test_utc_timezone_edge_cases():
+    """Test UTC timezone handling in edge cases."""
+    # Test with leap year
+    leap_date = datetime(2024, 2, 29, tzinfo=UTC)
+    assert leap_date.tzinfo == timezone.utc
+    
+    # Test DST transition dates
+    dst_date = datetime(2024, 3, 10, 2, 0, tzinfo=UTC)  # US DST start
+    assert dst_date.tzinfo == timezone.utc
+    
+    # Test with minimum/maximum dates
+    min_date = datetime.min.replace(tzinfo=UTC)
+    max_date = datetime.max.replace(tzinfo=UTC)
+    assert min_date.tzinfo == timezone.utc
+    assert max_date.tzinfo == timezone.utc
+    
+    # Test timezone offset calculations
+    dt = datetime(2024, 1, 1, tzinfo=UTC)
+    offset = dt.utcoffset()
+    assert offset == timedelta(0)  # UTC should always have zero offset

tests/utilities/test_events.py

@@ -0,0 +1,497 @@
+import json
+from datetime import datetime
+from unittest.mock import MagicMock, patch
+
+import pytest
+from pydantic import Field
+
+from crewai.agent import Agent
+from crewai.agents.crew_agent_executor import CrewAgentExecutor
+from crewai.crew import Crew
+from crewai.flow.flow import Flow, listen, start
+from crewai.task import Task
+from crewai.tools.base_tool import BaseTool
+from crewai.tools.tool_usage import ToolUsage
+from crewai.utilities.events.agent_events import (
+    AgentExecutionCompletedEvent,
+    AgentExecutionErrorEvent,
+    AgentExecutionStartedEvent,
+)
+from crewai.utilities.events.crew_events import (
+    CrewKickoffCompletedEvent,
+    CrewKickoffFailedEvent,
+    CrewKickoffStartedEvent,
+)
+from crewai.utilities.events.crewai_event_bus import crewai_event_bus
+from crewai.utilities.events.event_types import ToolUsageFinishedEvent
+from crewai.utilities.events.flow_events import (
+    FlowCreatedEvent,
+    FlowFinishedEvent,
+    FlowStartedEvent,
+    MethodExecutionFailedEvent,
+    MethodExecutionStartedEvent,
+)
+from crewai.utilities.events.task_events import (
+    TaskCompletedEvent,
+    TaskFailedEvent,
+    TaskStartedEvent,
+)
+from crewai.utilities.events.tool_usage_events import (
+    ToolUsageErrorEvent,
+)
+
+base_agent = Agent(
+    role="base_agent",
+    llm="gpt-4o-mini",
+    goal="Just say hi",
+    backstory="You are a helpful assistant that just says hi",
+)
+
+base_task = Task(
+    description="Just say hi",
+    expected_output="hi",
+    agent=base_agent,
+)
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_crew_emits_start_kickoff_event():
+    received_events = []
+
+    with crewai_event_bus.scoped_handlers():
+
+        @crewai_event_bus.on(CrewKickoffStartedEvent)
+        def handle_crew_start(source, event):
+            received_events.append(event)
+
+        crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+
+        crew.kickoff()
+
+        assert len(received_events) == 1
+        assert received_events[0].crew_name == "TestCrew"
+        assert isinstance(received_events[0].timestamp, datetime)
+        assert received_events[0].type == "crew_kickoff_started"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_crew_emits_end_kickoff_event():
+    received_events = []
+
+    @crewai_event_bus.on(CrewKickoffCompletedEvent)
+    def handle_crew_end(source, event):
+        received_events.append(event)
+
+    crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+
+    crew.kickoff()
+
+    assert len(received_events) == 1
+    assert received_events[0].crew_name == "TestCrew"
+    assert isinstance(received_events[0].timestamp, datetime)
+    assert received_events[0].type == "crew_kickoff_completed"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_crew_emits_kickoff_failed_event():
+    received_events = []
+
+    with crewai_event_bus.scoped_handlers():
+
+        @crewai_event_bus.on(CrewKickoffFailedEvent)
+        def handle_crew_failed(source, event):
+            received_events.append(event)
+
+        crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+
+        with patch.object(Crew, "_execute_tasks") as mock_execute:
+            error_message = "Simulated crew kickoff failure"
+            mock_execute.side_effect = Exception(error_message)
+
+            with pytest.raises(Exception):
+                crew.kickoff()
+
+        assert len(received_events) == 1
+        assert received_events[0].error == error_message
+        assert isinstance(received_events[0].timestamp, datetime)
+        assert received_events[0].type == "crew_kickoff_failed"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_crew_emits_start_task_event():
+    received_events = []
+
+    @crewai_event_bus.on(TaskStartedEvent)
+    def handle_task_start(source, event):
+        received_events.append(event)
+
+    crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+
+    crew.kickoff()
+
+    assert len(received_events) == 1
+    assert isinstance(received_events[0].timestamp, datetime)
+    assert received_events[0].type == "task_started"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_crew_emits_end_task_event():
+    received_events = []
+
+    @crewai_event_bus.on(TaskCompletedEvent)
+    def handle_task_end(source, event):
+        received_events.append(event)
+
+    crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+
+    crew.kickoff()
+
+    assert len(received_events) == 1
+    assert isinstance(received_events[0].timestamp, datetime)
+    assert received_events[0].type == "task_completed"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_task_emits_failed_event_on_execution_error():
+    received_events = []
+    received_sources = []
+
+    @crewai_event_bus.on(TaskFailedEvent)
+    def handle_task_failed(source, event):
+        received_events.append(event)
+        received_sources.append(source)
+
+    with patch.object(
+        Task,
+        "_execute_core",
+    ) as mock_execute:
+        error_message = "Simulated task failure"
+        mock_execute.side_effect = Exception(error_message)
+        agent = Agent(
+            role="base_agent",
+            goal="Just say hi",
+            backstory="You are a helpful assistant that just says hi",
+        )
+        task = Task(
+            description="Just say hi",
+            expected_output="hi",
+            agent=agent,
+        )
+
+        with pytest.raises(Exception):
+            agent.execute_task(task=task)
+
+            assert len(received_events) == 1
+            assert received_sources[0] == task
+            assert received_events[0].error == error_message
+            assert isinstance(received_events[0].timestamp, datetime)
+            assert received_events[0].type == "task_failed"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_agent_emits_execution_started_and_completed_events():
+    received_events = []
+
+    @crewai_event_bus.on(AgentExecutionStartedEvent)
+    def handle_agent_start(source, event):
+        received_events.append(event)
+
+    @crewai_event_bus.on(AgentExecutionCompletedEvent)
+    def handle_agent_completed(source, event):
+        received_events.append(event)
+
+    crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+    crew.kickoff()
+    assert len(received_events) == 2
+    assert received_events[0].agent == base_agent
+    assert received_events[0].task == base_task
+    assert received_events[0].tools == []
+    assert isinstance(received_events[0].task_prompt, str)
+    assert (
+        received_events[0].task_prompt
+        == "Just say hi\n\nThis is the expected criteria for your final answer: hi\nyou MUST return the actual complete content as the final answer, not a summary."
+    )
+    assert isinstance(received_events[0].timestamp, datetime)
+    assert received_events[0].type == "agent_execution_started"
+    assert isinstance(received_events[1].timestamp, datetime)
+    assert received_events[1].type == "agent_execution_completed"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_agent_emits_execution_error_event():
+    received_events = []
+
+    @crewai_event_bus.on(AgentExecutionErrorEvent)
+    def handle_agent_start(source, event):
+        received_events.append(event)
+
+    error_message = "Error happening while sending prompt to model."
+    base_agent.max_retry_limit = 0
+    with patch.object(
+        CrewAgentExecutor, "invoke", wraps=base_agent.agent_executor.invoke
+    ) as invoke_mock:
+        invoke_mock.side_effect = Exception(error_message)
+
+        with pytest.raises(Exception) as e:
+            base_agent.execute_task(
+                task=base_task,
+            )
+
+        assert len(received_events) == 1
+        assert received_events[0].agent == base_agent
+        assert received_events[0].task == base_task
+        assert received_events[0].error == error_message
+        assert isinstance(received_events[0].timestamp, datetime)
+        assert received_events[0].type == "agent_execution_error"
+
+
+class SayHiTool(BaseTool):
+    name: str = Field(default="say_hi", description="The name of the tool")
+    description: str = Field(
+        default="Say hi", description="The description of the tool"
+    )
+
+    def _run(self) -> str:
+        return "hi"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_tools_emits_finished_events():
+    received_events = []
+
+    @crewai_event_bus.on(ToolUsageFinishedEvent)
+    def handle_tool_end(source, event):
+        received_events.append(event)
+
+    agent = Agent(
+        role="base_agent",
+        goal="Just say hi",
+        backstory="You are a helpful assistant that just says hi",
+        tools=[SayHiTool()],
+    )
+
+    task = Task(
+        description="Just say hi",
+        expected_output="hi",
+        agent=agent,
+    )
+    crew = Crew(agents=[agent], tasks=[task], name="TestCrew")
+    crew.kickoff()
+    assert len(received_events) == 1
+    assert received_events[0].agent_key == agent.key
+    assert received_events[0].agent_role == agent.role
+    assert received_events[0].tool_name == SayHiTool().name
+    assert received_events[0].tool_args == {}
+    assert received_events[0].type == "tool_usage_finished"
+    assert isinstance(received_events[0].timestamp, datetime)
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_tools_emits_error_events():
+    received_events = []
+
+    @crewai_event_bus.on(ToolUsageErrorEvent)
+    def handle_tool_end(source, event):
+        received_events.append(event)
+
+    class ErrorTool(BaseTool):
+        name: str = Field(
+            default="error_tool", description="A tool that raises an error"
+        )
+        description: str = Field(
+            default="This tool always raises an error",
+            description="The description of the tool",
+        )
+
+        def _run(self) -> str:
+            raise Exception("Simulated tool error")
+
+    agent = Agent(
+        role="base_agent",
+        goal="Try to use the error tool",
+        backstory="You are an assistant that tests error handling",
+        tools=[ErrorTool()],
+    )
+
+    task = Task(
+        description="Use the error tool",
+        expected_output="This should error",
+        agent=agent,
+    )
+
+    crew = Crew(agents=[agent], tasks=[task], name="TestCrew")
+    crew.kickoff()
+
+    assert len(received_events) == 75
+    assert received_events[0].agent_key == agent.key
+    assert received_events[0].agent_role == agent.role
+    assert received_events[0].tool_name == "error_tool"
+    assert received_events[0].tool_args == {}
+    assert str(received_events[0].error) == "Simulated tool error"
+    assert received_events[0].type == "tool_usage_error"
+    assert isinstance(received_events[0].timestamp, datetime)
+
+
+def test_flow_emits_start_event():
+    received_events = []
+
+    with crewai_event_bus.scoped_handlers():
+
+        @crewai_event_bus.on(FlowStartedEvent)
+        def handle_flow_start(source, event):
+            received_events.append(event)
+
+        class TestFlow(Flow[dict]):
+            @start()
+            def begin(self):
+                return "started"
+
+        flow = TestFlow()
+        flow.kickoff()
+
+        assert len(received_events) == 1
+        assert received_events[0].flow_name == "TestFlow"
+        assert received_events[0].type == "flow_started"
+
+
+def test_flow_emits_finish_event():
+    received_events = []
+
+    with crewai_event_bus.scoped_handlers():
+
+        @crewai_event_bus.on(FlowFinishedEvent)
+        def handle_flow_finish(source, event):
+            received_events.append(event)
+
+        class TestFlow(Flow[dict]):
+            @start()
+            def begin(self):
+                return "completed"
+
+        flow = TestFlow()
+        result = flow.kickoff()
+
+        assert len(received_events) == 1
+        assert received_events[0].flow_name == "TestFlow"
+        assert received_events[0].type == "flow_finished"
+        assert received_events[0].result == "completed"
+        assert result == "completed"
+
+
+def test_flow_emits_method_execution_started_event():
+    received_events = []
+
+    with crewai_event_bus.scoped_handlers():
+
+        @crewai_event_bus.on(MethodExecutionStartedEvent)
+        def handle_method_start(source, event):
+            print("event in method name", event.method_name)
+            received_events.append(event)
+
+        class TestFlow(Flow[dict]):
+            @start()
+            def begin(self):
+                return "started"
+
+            @listen("begin")
+            def second_method(self):
+                return "executed"
+
+        flow = TestFlow()
+        flow.kickoff()
+
+        assert len(received_events) == 2
+
+        assert received_events[0].method_name == "begin"
+        assert received_events[0].flow_name == "TestFlow"
+        assert received_events[0].type == "method_execution_started"
+
+        assert received_events[1].method_name == "second_method"
+        assert received_events[1].flow_name == "TestFlow"
+        assert received_events[1].type == "method_execution_started"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_register_handler_adds_new_handler():
+    received_events = []
+
+    def custom_handler(source, event):
+        received_events.append(event)
+
+    with crewai_event_bus.scoped_handlers():
+        crewai_event_bus.register_handler(CrewKickoffStartedEvent, custom_handler)
+
+        crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+        crew.kickoff()
+
+        assert len(received_events) == 1
+        assert isinstance(received_events[0].timestamp, datetime)
+        assert received_events[0].type == "crew_kickoff_started"
+
+
+@pytest.mark.vcr(filter_headers=["authorization"])
+def test_multiple_handlers_for_same_event():
+    received_events_1 = []
+    received_events_2 = []
+
+    def handler_1(source, event):
+        received_events_1.append(event)
+
+    def handler_2(source, event):
+        received_events_2.append(event)
+
+    with crewai_event_bus.scoped_handlers():
+        crewai_event_bus.register_handler(CrewKickoffStartedEvent, handler_1)
+        crewai_event_bus.register_handler(CrewKickoffStartedEvent, handler_2)
+
+        crew = Crew(agents=[base_agent], tasks=[base_task], name="TestCrew")
+        crew.kickoff()
+
+        assert len(received_events_1) == 1
+        assert len(received_events_2) == 1
+        assert received_events_1[0].type == "crew_kickoff_started"
+        assert received_events_2[0].type == "crew_kickoff_started"
+
+
+def test_flow_emits_created_event():
+    received_events = []
+
+    @crewai_event_bus.on(FlowCreatedEvent)
+    def handle_flow_created(source, event):
+        received_events.append(event)
+
+    class TestFlow(Flow[dict]):
+        @start()
+        def begin(self):
+            return "started"
+
+    flow = TestFlow()
+    flow.kickoff()
+
+    assert len(received_events) == 1
+    assert received_events[0].flow_name == "TestFlow"
+    assert received_events[0].type == "flow_created"
+
+
+def test_flow_emits_method_execution_failed_event():
+    received_events = []
+    error = Exception("Simulated method failure")
+
+    @crewai_event_bus.on(MethodExecutionFailedEvent)
+    def handle_method_failed(source, event):
+        received_events.append(event)
+
+    class TestFlow(Flow[dict]):
+        @start()
+        def begin(self):
+            raise error
+
+    flow = TestFlow()
+    with pytest.raises(Exception):
+        flow.kickoff()
+
+    assert len(received_events) == 1
+    assert received_events[0].method_name == "begin"
+    assert received_events[0].flow_name == "TestFlow"
+    assert received_events[0].type == "method_execution_failed"
+    assert received_events[0].error == error