refactor: incorporate improvements from PR #2178 into guardrail validation

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

.gitignore

@@ -21,5 +21,4 @@ crew_tasks_output.json
 .mypy_cache
 .ruff_cache
 .venv
-agentops.log
-test_flow.html
+agentops.log

docs/how-to/langfuse-observability.mdx

@@ -10,8 +10,6 @@ This notebook demonstrates how to integrate **Langfuse** with **CrewAI** using O
 
 > **What is Langfuse?** [Langfuse](https://langfuse.com) is an open-source LLM engineering platform. It provides tracing and monitoring capabilities for LLM applications, helping developers debug, analyze, and optimize their AI systems. Langfuse integrates with various tools and frameworks via native integrations, OpenTelemetry, and APIs/SDKs.
 
-[![Langfuse Overview Video](https://github.com/user-attachments/assets/3926b288-ff61-4b95-8aa1-45d041c70866)](https://langfuse.com/watch-demo)
-
 ## Get Started
 
 We'll walk through a simple example of using CrewAI and integrating it with Langfuse via OpenTelemetry using OpenLit.

src/crewai/agents/agent_builder/base_agent_executor_mixin.py

@@ -114,15 +114,10 @@ class CrewAgentExecutorMixin:
             prompt = (
                 "\n\n=====\n"
                 "## HUMAN FEEDBACK: Provide feedback on the Final Result and Agent's actions.\n"
-                "Please follow these guidelines:\n"
-                " - If you are happy with the result, simply hit Enter without typing anything.\n"
-                " - Otherwise, provide specific improvement requests.\n"
-                " - You can provide multiple rounds of feedback until satisfied.\n"
+                "Respond with 'looks good' to accept or provide specific improvement requests.\n"
+                "You can provide multiple rounds of feedback until satisfied.\n"
                 "=====\n"
             )
 
         self._printer.print(content=prompt, color="bold_yellow")
-        response = input()
-        if response.strip() != "":
-            self._printer.print(content="\nProcessing your feedback...", color="cyan")
-        return response
+        return input()

src/crewai/agents/agent_builder/utilities/base_output_converter.py

@@ -31,11 +31,11 @@ class OutputConverter(BaseModel, ABC):
     )
 
     @abstractmethod
-    def to_pydantic(self, current_attempt=1) -> BaseModel:
+    def to_pydantic(self, current_attempt=1):
         """Convert text to pydantic."""
         pass
 
     @abstractmethod
-    def to_json(self, current_attempt=1) -> dict:
+    def to_json(self, current_attempt=1):
         """Convert text to json."""
         pass

src/crewai/agents/agent_builder/utilities/base_token_process.py

@@ -1,9 +1,5 @@
-import logging
-
 from crewai.types.usage_metrics import UsageMetrics
 
-logger = logging.getLogger(__name__)
-
 
 class TokenProcess:
     def __init__(self) -> None:
@@ -21,21 +17,7 @@ class TokenProcess:
         self.completion_tokens += tokens
         self.total_tokens += tokens
 
-    def sum_cached_prompt_tokens(self, tokens: int | None) -> None:
-        """
-        Adds the given token count to cached prompt tokens.
-        
-        Args:
-            tokens (int | None): Number of tokens to add. None values are ignored.
-            
-        Raises:
-            ValueError: If tokens is negative.
-        """
-        if tokens is None:
-            logger.debug("Received None value for token count")
-            return
-        if tokens < 0:
-            raise ValueError("Token count cannot be negative")
+    def sum_cached_prompt_tokens(self, tokens: int) -> None:
         self.cached_prompt_tokens += tokens
 
     def sum_successful_requests(self, requests: int) -> None:

src/crewai/agents/crew_agent_executor.py

@@ -548,6 +548,10 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
         self, initial_answer: AgentFinish, feedback: str
     ) -> AgentFinish:
         """Process feedback for training scenarios with single iteration."""
+        self._printer.print(
+            content="\nProcessing training feedback.\n",
+            color="yellow",
+        )
         self._handle_crew_training_output(initial_answer, feedback)
         self.messages.append(
             self._format_msg(
@@ -567,8 +571,9 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
         answer = current_answer
 
         while self.ask_for_human_input:
-            # If the user provides a blank response, assume they are happy with the result
-            if feedback.strip() == "":
+            response = self._get_llm_feedback_response(feedback)
+
+            if not self._feedback_requires_changes(response):
                 self.ask_for_human_input = False
             else:
                 answer = self._process_feedback_iteration(feedback)
@@ -576,6 +581,27 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
 
         return answer
 
+    def _get_llm_feedback_response(self, feedback: str) -> Optional[str]:
+        """Get LLM classification of whether feedback requires changes."""
+        prompt = self._i18n.slice("human_feedback_classification").format(
+            feedback=feedback
+        )
+        message = self._format_msg(prompt, role="system")
+
+        for retry in range(MAX_LLM_RETRY):
+            try:
+                response = self.llm.call([message], callbacks=self.callbacks)
+                return response.strip().lower() if response else None
+            except Exception as error:
+                self._log_feedback_error(retry, error)
+
+        self._log_max_retries_exceeded()
+        return None
+
+    def _feedback_requires_changes(self, response: Optional[str]) -> bool:
+        """Determine if feedback response indicates need for changes."""
+        return response == "true" if response else False
+
     def _process_feedback_iteration(self, feedback: str) -> AgentFinish:
         """Process a single feedback iteration."""
         self.messages.append(

src/crewai/llm.py

@@ -26,9 +26,9 @@ from crewai.utilities.events.tool_usage_events import ToolExecutionErrorEvent
 with warnings.catch_warnings():
     warnings.simplefilter("ignore", UserWarning)
     import litellm
-    from litellm import Choices
+    from litellm import Choices, get_supported_openai_params
     from litellm.types.utils import ModelResponse
-    from litellm.utils import get_supported_openai_params, supports_response_schema
+    from litellm.utils import supports_response_schema
 
 
 from crewai.traces.unified_trace_controller import trace_llm_call
@@ -449,7 +449,7 @@ class LLM:
     def supports_function_calling(self) -> bool:
         try:
             params = get_supported_openai_params(model=self.model)
-            return params is not None and "tools" in params
+            return "response_format" in params
         except Exception as e:
             logging.error(f"Failed to get supported params: {str(e)}")
             return False
@@ -457,7 +457,7 @@ class LLM:
     def supports_stop_words(self) -> bool:
         try:
             params = get_supported_openai_params(model=self.model)
-            return params is not None and "stop" in params
+            return "stop" in params
         except Exception as e:
             logging.error(f"Failed to get supported params: {str(e)}")
             return False

src/crewai/task.py

@@ -3,17 +3,20 @@ import inspect
 import json
 import logging
 import threading
+import typing
 import uuid
 from concurrent.futures import Future
 from copy import copy
 from hashlib import md5
 from pathlib import Path
 from typing import (
+    AbstractSet,
     Any,
     Callable,
     ClassVar,
     Dict,
     List,
+    Mapping,
     Optional,
     Set,
     Tuple,
@@ -32,6 +35,7 @@ from pydantic import (
 from pydantic_core import PydanticCustomError
 
 from crewai.agents.agent_builder.base_agent import BaseAgent
+from crewai.tasks.exceptions import GuardrailValidationError
 from crewai.tasks.guardrail_result import GuardrailResult
 from crewai.tasks.output_format import OutputFormat
 from crewai.tasks.task_output import TaskOutput
@@ -113,7 +117,7 @@ class Task(BaseModel):
         description="Task output, it's final result after being executed", default=None
     )
     tools: Optional[List[BaseTool]] = Field(
-        default_factory=list,
+        default_factory=list[BaseTool],
         description="Tools the agent is limited to use for this task.",
     )
     id: UUID4 = Field(
@@ -129,7 +133,7 @@ class Task(BaseModel):
         description="A converter class used to export structured output",
         default=None,
     )
-    processed_by_agents: Set[str] = Field(default_factory=set)
+    processed_by_agents: Set[str] = Field(default_factory=set[str])
     guardrail: Optional[Callable[[TaskOutput], Tuple[bool, Any]]] = Field(
         default=None,
         description="Function to validate task output before proceeding to next task",
@@ -151,8 +155,8 @@ class Task(BaseModel):
         """Validate that the guardrail function has the correct signature and behavior.
 
         While type hints provide static checking, this validator ensures runtime safety by:
-        1. Verifying the function accepts exactly one parameter (the TaskOutput)
-        2. Checking return type annotations match Tuple[bool, Any] if present
+        1. Verifying the function accepts exactly one required positional parameter (the TaskOutput)
+        2. Checking return type annotations match tuple[bool, Any] or specific types like tuple[bool, str]
         3. Providing clear, immediate error messages for debugging
 
         This runtime validation is crucial because:
@@ -160,6 +164,24 @@ class Task(BaseModel):
         - Function signatures need immediate validation before task execution
         - Clear error messages help users debug guardrail implementation issues
 
+        Examples:
+            Simple validation with new style annotation:
+            >>> def validate_output(result: TaskOutput) -> tuple[bool, str]:
+            ...     return (True, result.raw.upper())
+
+            Validation with optional parameters:
+            >>> def validate_with_options(result: TaskOutput, strict: bool = True) -> tuple[bool, str]:
+            ...     if strict and not result.raw.isupper():
+            ...         return (False, "Text must be uppercase")
+            ...     return (True, result.raw)
+
+            Validation with specific return type:
+            >>> def validate_task_output(result: TaskOutput) -> tuple[bool, TaskOutput]:
+            ...     if not result.raw:
+            ...         return (False, result)
+            ...     result.raw = result.raw.strip()
+            ...     return (True, result)
+
         Args:
             v: The guardrail function to validate
 
@@ -168,22 +190,46 @@ class Task(BaseModel):
 
         Raises:
             ValueError: If the function signature is invalid or return annotation
-                       doesn't match Tuple[bool, Any]
+                       doesn't match tuple[bool, Any] or specific allowed types
         """
         if v is not None:
             sig = inspect.signature(v)
-            if len(sig.parameters) != 1:
-                raise ValueError("Guardrail function must accept exactly one parameter")
+            # Check for exactly one required positional parameter
+            positional_args = [
+                param for param in sig.parameters.values()
+                if param.default is inspect.Parameter.empty
+            ]
+            if len(positional_args) != 1:
+                raise GuardrailValidationError(
+                    "Guardrail function must accept exactly one required parameter",
+                    {"params": [str(p) for p in sig.parameters.values()]}
+                )
 
             # Check return annotation if present, but don't require it
-            return_annotation = sig.return_annotation
-            if return_annotation != inspect.Signature.empty:
+            type_hints = typing.get_type_hints(v)
+            return_annotation = type_hints.get('return')
+            if return_annotation:
+                # Simplified type checking logic
+                return_annotation_args = typing.get_args(return_annotation)
                 if not (
-                    return_annotation == Tuple[bool, Any]
-                    or str(return_annotation) == "Tuple[bool, Any]"
+                    typing.get_origin(return_annotation) is tuple
+                    and len(return_annotation_args) == 2
+                    and return_annotation_args[0] is bool
+                    and (
+                        return_annotation_args[1] is Any
+                        or return_annotation_args[1] is str
+                        or return_annotation_args[1] is TaskOutput
+                        or return_annotation_args[1] == Union[str, TaskOutput]
+                        or (
+                            typing.get_origin(return_annotation_args[1]) is dict
+                            and typing.get_args(return_annotation_args[1])[0] is str
+                            and typing.get_args(return_annotation_args[1])[1] is Any
+                        )
+                    )
                 ):
-                    raise ValueError(
-                        "If return type is annotated, it must be Tuple[bool, Any]"
+                    raise GuardrailValidationError(
+                        "Invalid return type annotation. Expected Tuple[bool, Any|str|TaskOutput|Dict[str, Any]]",
+                        {"got": str(return_annotation)}
                     )
         return v
 
@@ -411,6 +457,7 @@ class Task(BaseModel):
                         "Task guardrail returned None as result. This is not allowed."
                     )
 
+                # Handle different result types
                 if isinstance(guardrail_result.result, str):
                     task_output.raw = guardrail_result.result
                     pydantic_output, json_output = self._export_output(
@@ -420,6 +467,13 @@ class Task(BaseModel):
                     task_output.json_dict = json_output
                 elif isinstance(guardrail_result.result, TaskOutput):
                     task_output = guardrail_result.result
+                elif isinstance(guardrail_result.result, dict):
+                    task_output.raw = guardrail_result.result
+                    task_output.json_dict = guardrail_result.result
+                    pydantic_output, _ = self._export_output(
+                        json.dumps(guardrail_result.result)
+                    )
+                    task_output.pydantic = pydantic_output
 
             self.output = task_output
             self.end_time = datetime.datetime.now()
@@ -610,40 +664,74 @@ class Task(BaseModel):
         self.delegations += 1
 
     def copy(
-        self, agents: List["BaseAgent"], task_mapping: Dict[str, "Task"]
+        self,
+        agents: List["BaseAgent"] | None = None,
+        task_mapping: Dict[str, "Task"] | None = None,
+        *,
+        include: AbstractSet[int] | AbstractSet[str] | Mapping[int, Any] | Mapping[str, Any] | None = None,
+        exclude: AbstractSet[int] | AbstractSet[str] | Mapping[int, Any] | Mapping[str, Any] | None = None,
+        update: dict[str, Any] | None = None,
+        deep: bool = False,
     ) -> "Task":
-        """Create a deep copy of the Task."""
-        exclude = {
+        """Create a deep copy of the Task.
+        
+        Args:
+            agents: Optional list of agents to copy agent references
+            task_mapping: Optional mapping of task keys to tasks for context
+            include: Fields to include in the copy
+            exclude: Fields to exclude from the copy
+            update: Fields to update in the copy
+            deep: Whether to perform a deep copy
+        """
+        if agents is None and task_mapping is None:
+            # New style copy using BaseModel
+            copied = super().copy(
+                include=include,
+                exclude=exclude,
+                update=update,
+                deep=deep,
+            )
+            
+            # Copy mutable fields
+            if self.tools:
+                copied.tools = copy(self.tools)
+            if self.context:
+                copied.context = copy(self.context)
+                
+            return copied
+            
+        # Legacy copy behavior
+        exclude_fields = {
             "id",
             "agent",
             "context",
             "tools",
         }
 
-        copied_data = self.model_dump(exclude=exclude)
+        copied_data = self.model_dump(exclude=exclude_fields)
         copied_data = {k: v for k, v in copied_data.items() if v is not None}
 
         cloned_context = (
             [task_mapping[context_task.key] for context_task in self.context]
-            if self.context
+            if self.context and task_mapping
             else None
         )
 
         def get_agent_by_role(role: str) -> Union["BaseAgent", None]:
+            if not agents:
+                return None
             return next((agent for agent in agents if agent.role == role), None)
 
         cloned_agent = get_agent_by_role(self.agent.role) if self.agent else None
         cloned_tools = copy(self.tools) if self.tools else []
 
-        copied_task = Task(
+        return Task(
             **copied_data,
             context=cloned_context,
             agent=cloned_agent,
             tools=cloned_tools,
         )
 
-        return copied_task
-
     def _export_output(
         self, result: str
     ) -> Tuple[Optional[BaseModel], Optional[Dict[str, Any]]]:

src/crewai/tasks/exceptions.py

@@ -0,0 +1,25 @@
+"""
+Module for task-related exceptions.
+
+This module provides custom exceptions used throughout the task system
+to provide more specific error handling and context.
+"""
+
+from typing import Any, Dict, Optional
+
+
+class GuardrailValidationError(Exception):
+    """Exception raised for guardrail validation errors.
+    
+    This exception provides detailed context about why a guardrail
+    validation failed, including the specific validation that failed
+    and any relevant context information.
+
+    Attributes:
+        message: A clear description of the validation error
+        context: Optional dictionary containing additional error context
+    """
+    def __init__(self, message: str, context: Optional[Dict[str, Any]] = None):
+        self.message = message
+        self.context = context or {}
+        super().__init__(self.message)

src/crewai/tasks/task_output.py

@@ -1,5 +1,5 @@
 import json
-from typing import Any, Dict, Optional
+from typing import Any, Dict, Optional, Union
 
 from pydantic import BaseModel, Field, model_validator
 
@@ -15,7 +15,7 @@ class TaskOutput(BaseModel):
         description="Expected output of the task", default=None
     )
     summary: Optional[str] = Field(description="Summary of the task", default=None)
-    raw: str = Field(description="Raw output of the task", default="")
+    raw: Any = Field(description="Raw output of the task", default="")
     pydantic: Optional[BaseModel] = Field(
         description="Pydantic output of task", default=None
     )

src/crewai/translations/en.json

@@ -23,6 +23,7 @@
     "summary": "This is a summary of our conversation so far:\n{merged_summary}",
     "manager_request": "Your best answer to your coworker asking you this, accounting for the context shared.",
     "formatted_task_instructions": "Ensure your final answer contains only the content in the following format: {output_format}\n\nEnsure the final output does not include any code block markers like ```json or ```python.",
+    "human_feedback_classification": "Determine if the following feedback indicates that the user is satisfied or if further changes are needed. Respond with 'True' if further changes are needed, or 'False' if the user is satisfied. **Important** Do not include any additional commentary outside of your 'True' or 'False' response.\n\nFeedback: \"{feedback}\"",
     "conversation_history_instruction": "You are a member of a crew collaborating to achieve a common goal. Your task is a specific action that contributes to this larger objective. For additional context, please review the conversation history between you and the user that led to the initiation of this crew. Use any relevant information or feedback from the conversation to inform your task execution and ensure your response aligns with both the immediate task and the crew's overall goals.",
     "feedback_instructions": "User feedback: {feedback}\nInstructions: Use this feedback to enhance the next output iteration.\nNote: Do not respond or add commentary."
   },

src/crewai/utilities/converter.py

@@ -20,11 +20,11 @@ class ConverterError(Exception):
 class Converter(OutputConverter):
     """Class that converts text into either pydantic or json."""
 
-    def to_pydantic(self, current_attempt=1) -> BaseModel:
+    def to_pydantic(self, current_attempt=1):
         """Convert text to pydantic."""
         try:
             if self.llm.supports_function_calling():
-                result = self._create_instructor().to_pydantic()
+                return self._create_instructor().to_pydantic()
             else:
                 response = self.llm.call(
                     [
@@ -32,40 +32,18 @@ class Converter(OutputConverter):
                         {"role": "user", "content": self.text},
                     ]
                 )
-                try:
-                    # Try to directly validate the response JSON
-                    result = self.model.model_validate_json(response)
-                except ValidationError:
-                    # If direct validation fails, attempt to extract valid JSON
-                    result = handle_partial_json(response, self.model, False, None)
-                    # Ensure result is a BaseModel instance
-                    if not isinstance(result, BaseModel):
-                        if isinstance(result, dict):
-                            result = self.model.parse_obj(result)
-                        elif isinstance(result, str):
-                            try:
-                                parsed = json.loads(result)
-                                result = self.model.parse_obj(parsed)
-                            except Exception as parse_err:
-                                raise ConverterError(
-                                    f"Failed to convert partial JSON result into Pydantic: {parse_err}"
-                                )
-                        else:
-                            raise ConverterError(
-                                "handle_partial_json returned an unexpected type."
-                            )
-            return result
+                return self.model.model_validate_json(response)
         except ValidationError as e:
             if current_attempt < self.max_attempts:
                 return self.to_pydantic(current_attempt + 1)
             raise ConverterError(
-                f"Failed to convert text into a Pydantic model due to validation error: {e}"
+                f"Failed to convert text into a Pydantic model due to the following validation error: {e}"
             )
         except Exception as e:
             if current_attempt < self.max_attempts:
                 return self.to_pydantic(current_attempt + 1)
             raise ConverterError(
-                f"Failed to convert text into a Pydantic model due to error: {e}"
+                f"Failed to convert text into a Pydantic model due to the following error: {e}"
             )
 
     def to_json(self, current_attempt=1):
@@ -219,15 +197,11 @@ def get_conversion_instructions(model: Type[BaseModel], llm: Any) -> str:
     if llm.supports_function_calling():
         model_schema = PydanticSchemaParser(model=model).get_schema()
         instructions += (
-            f"\n\nOutput ONLY the valid JSON and nothing else.\n\n"
-            f"The JSON must follow this schema exactly:\n```json\n{model_schema}\n```"
+            f"\n\nThe JSON should follow this schema:\n```json\n{model_schema}\n```"
         )
     else:
         model_description = generate_model_description(model)
-        instructions += (
-            f"\n\nOutput ONLY the valid JSON and nothing else.\n\n"
-            f"The JSON must follow this format exactly:\n{model_description}"
-        )
+        instructions += f"\n\nThe JSON should follow this format:\n{model_description}"
     return instructions
 
 

tests/agent_test.py

@@ -1,6 +1,7 @@
 """Test Agent creation and execution basic functionality."""
 
 import os
+from datetime import UTC, datetime, timezone
 from unittest import mock
 from unittest.mock import patch
 
@@ -8,7 +9,7 @@ import pytest
 
 from crewai import Agent, Crew, Task
 from crewai.agents.cache import CacheHandler
-from crewai.agents.crew_agent_executor import AgentFinish, CrewAgentExecutor
+from crewai.agents.crew_agent_executor import CrewAgentExecutor
 from crewai.agents.parser import AgentAction, CrewAgentParser, OutputParserException
 from crewai.knowledge.source.base_knowledge_source import BaseKnowledgeSource
 from crewai.knowledge.source.string_knowledge_source import StringKnowledgeSource
@@ -998,35 +999,23 @@ def test_agent_human_input():
     # Side effect function for _ask_human_input to simulate multiple feedback iterations
     feedback_responses = iter(
         [
-            "Don't say hi, say Hello instead!",  # First feedback: instruct change
-            "",  # Second feedback: empty string signals acceptance
+            "Don't say hi, say Hello instead!",  # First feedback
+            "looks good",  # Second feedback to exit loop
         ]
     )
 
     def ask_human_input_side_effect(*args, **kwargs):
         return next(feedback_responses)
 
-    # Patch both _ask_human_input and _invoke_loop to avoid real API/network calls.
-    with (
-        patch.object(
-            CrewAgentExecutor,
-            "_ask_human_input",
-            side_effect=ask_human_input_side_effect,
-        ) as mock_human_input,
-        patch.object(
-            CrewAgentExecutor,
-            "_invoke_loop",
-            return_value=AgentFinish(output="Hello", thought="", text=""),
-        ) as mock_invoke_loop,
-    ):
+    with patch.object(
+        CrewAgentExecutor, "_ask_human_input", side_effect=ask_human_input_side_effect
+    ) as mock_human_input:
         # Execute the task
         output = agent.execute_task(task)
 
-        # Assertions to ensure the agent behaves correctly.
-        # It should have requested feedback twice.
-        assert mock_human_input.call_count == 2
-        # The final result should be processed to "Hello"
-        assert output.strip().lower() == "hello"
+        # Assertions to ensure the agent behaves correctly
+        assert mock_human_input.call_count == 2  # Should have asked for feedback twice
+        assert output.strip().lower() == "hello"  # Final output should be 'Hello'
 
 
 def test_interpolate_inputs():

tests/cassettes/test_agent_human_input.yaml

@@ -0,0 +1,520 @@
+interactions:
+- request:
+    body: !!binary |
+      CqcXCiQKIgoMc2VydmljZS5uYW1lEhIKEGNyZXdBSS10ZWxlbWV0cnkS/hYKEgoQY3Jld2FpLnRl
+      bGVtZXRyeRJ5ChBuJJtOdNaB05mOW/p3915eEgj2tkAd3rZcASoQVG9vbCBVc2FnZSBFcnJvcjAB
+      OYa7/URvKBUYQUpcFEVvKBUYShoKDmNyZXdhaV92ZXJzaW9uEggKBjAuODYuMEoPCgNsbG0SCAoG
+      Z3B0LTRvegIYAYUBAAEAABLJBwoQifhX01E5i+5laGdALAlZBBIIBuGM1aN+OPgqDENyZXcgQ3Jl
+      YXRlZDABORVGruBvKBUYQaipwOBvKBUYShoKDmNyZXdhaV92ZXJzaW9uEggKBjAuODYuMEoaCg5w
+      eXRob25fdmVyc2lvbhIICgYzLjEyLjdKLgoIY3Jld19rZXkSIgogN2U2NjA4OTg5ODU5YTY3ZWVj
+      ODhlZWY3ZmNlODUyMjVKMQoHY3Jld19pZBImCiRiOThiNWEwMC01YTI1LTQxMDctYjQwNS1hYmYz
+      MjBhOGYzYThKHAoMY3Jld19wcm9jZXNzEgwKCnNlcXVlbnRpYWxKEQoLY3Jld19tZW1vcnkSAhAA
+      ShoKFGNyZXdfbnVtYmVyX29mX3Rhc2tzEgIYAUobChVjcmV3X251bWJlcl9vZl9hZ2VudHMSAhgB
+      SuQCCgtjcmV3X2FnZW50cxLUAgrRAlt7ImtleSI6ICIyMmFjZDYxMWU0NGVmNWZhYzA1YjUzM2Q3
+      NWU4ODkzYiIsICJpZCI6ICJkNWIyMzM1YS0yMmIyLTQyZWEtYmYwNS03OTc3NmU3MmYzOTIiLCAi
+      cm9sZSI6ICJEYXRhIFNjaWVudGlzdCIsICJ2ZXJib3NlPyI6IGZhbHNlLCAibWF4X2l0ZXIiOiAy
+      MCwgIm1heF9ycG0iOiBudWxsLCAiZnVuY3Rpb25fY2FsbGluZ19sbG0iOiAiIiwgImxsbSI6ICJn
+      cHQtNG8tbWluaSIsICJkZWxlZ2F0aW9uX2VuYWJsZWQ/IjogZmFsc2UsICJhbGxvd19jb2RlX2V4
+      ZWN1dGlvbj8iOiBmYWxzZSwgIm1heF9yZXRyeV9saW1pdCI6IDIsICJ0b29sc19uYW1lcyI6IFsi
+      Z2V0IGdyZWV0aW5ncyJdfV1KkgIKCmNyZXdfdGFza3MSgwIKgAJbeyJrZXkiOiAiYTI3N2IzNGIy
+      YzE0NmYwYzU2YzVlMTM1NmU4ZjhhNTciLCAiaWQiOiAiMjJiZWMyMzEtY2QyMS00YzU4LTgyN2Ut
+      MDU4MWE4ZjBjMTExIiwgImFzeW5jX2V4ZWN1dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lucHV0PyI6
+      IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJEYXRhIFNjaWVudGlzdCIsICJhZ2VudF9rZXkiOiAiMjJh
+      Y2Q2MTFlNDRlZjVmYWMwNWI1MzNkNzVlODg5M2IiLCAidG9vbHNfbmFtZXMiOiBbImdldCBncmVl
+      dGluZ3MiXX1degIYAYUBAAEAABKOAgoQ5WYoxRtTyPjge4BduhL0rRIIv2U6rvWALfwqDFRhc2sg
+      Q3JlYXRlZDABOX068uBvKBUYQZkv8+BvKBUYSi4KCGNyZXdfa2V5EiIKIDdlNjYwODk4OTg1OWE2
+      N2VlYzg4ZWVmN2ZjZTg1MjI1SjEKB2NyZXdfaWQSJgokYjk4YjVhMDAtNWEyNS00MTA3LWI0MDUt
+      YWJmMzIwYThmM2E4Si4KCHRhc2tfa2V5EiIKIGEyNzdiMzRiMmMxNDZmMGM1NmM1ZTEzNTZlOGY4
+      YTU3SjEKB3Rhc2tfaWQSJgokMjJiZWMyMzEtY2QyMS00YzU4LTgyN2UtMDU4MWE4ZjBjMTExegIY
+      AYUBAAEAABKQAQoQXyeDtJDFnyp2Fjk9YEGTpxIIaNE7gbhPNYcqClRvb2wgVXNhZ2UwATkaXTvj
+      bygVGEGvx0rjbygVGEoaCg5jcmV3YWlfdmVyc2lvbhIICgYwLjg2LjBKHAoJdG9vbF9uYW1lEg8K
+      DUdldCBHcmVldGluZ3NKDgoIYXR0ZW1wdHMSAhgBegIYAYUBAAEAABLVBwoQMWfznt0qwauEzl7T
+      UOQxRBII9q+pUS5EdLAqDENyZXcgQ3JlYXRlZDABORONPORvKBUYQSAoS+RvKBUYShoKDmNyZXdh
+      aV92ZXJzaW9uEggKBjAuODYuMEoaCg5weXRob25fdmVyc2lvbhIICgYzLjEyLjdKLgoIY3Jld19r
+      ZXkSIgogYzMwNzYwMDkzMjY3NjE0NDRkNTdjNzFkMWRhM2YyN2NKMQoHY3Jld19pZBImCiQ3OTQw
+      MTkyNS1iOGU5LTQ3MDgtODUzMC00NDhhZmEzYmY4YjBKHAoMY3Jld19wcm9jZXNzEgwKCnNlcXVl
+      bnRpYWxKEQoLY3Jld19tZW1vcnkSAhAAShoKFGNyZXdfbnVtYmVyX29mX3Rhc2tzEgIYAUobChVj
+      cmV3X251bWJlcl9vZl9hZ2VudHMSAhgBSuoCCgtjcmV3X2FnZW50cxLaAgrXAlt7ImtleSI6ICI5
+      OGYzYjFkNDdjZTk2OWNmMDU3NzI3Yjc4NDE0MjVjZCIsICJpZCI6ICI5OTJkZjYyZi1kY2FiLTQy
+      OTUtOTIwNi05MDBkNDExNGIxZTkiLCAicm9sZSI6ICJGcmllbmRseSBOZWlnaGJvciIsICJ2ZXJi
+      b3NlPyI6IGZhbHNlLCAibWF4X2l0ZXIiOiAyMCwgIm1heF9ycG0iOiBudWxsLCAiZnVuY3Rpb25f
+      Y2FsbGluZ19sbG0iOiAiIiwgImxsbSI6ICJncHQtNG8tbWluaSIsICJkZWxlZ2F0aW9uX2VuYWJs
+      ZWQ/IjogZmFsc2UsICJhbGxvd19jb2RlX2V4ZWN1dGlvbj8iOiBmYWxzZSwgIm1heF9yZXRyeV9s
+      aW1pdCI6IDIsICJ0b29sc19uYW1lcyI6IFsiZGVjaWRlIGdyZWV0aW5ncyJdfV1KmAIKCmNyZXdf
+      dGFza3MSiQIKhgJbeyJrZXkiOiAiODBkN2JjZDQ5MDk5MjkwMDgzODMyZjBlOTgzMzgwZGYiLCAi
+      aWQiOiAiMmZmNjE5N2UtYmEyNy00YjczLWI0YTctNGZhMDQ4ZTYyYjQ3IiwgImFzeW5jX2V4ZWN1
+      dGlvbj8iOiBmYWxzZSwgImh1bWFuX2lucHV0PyI6IGZhbHNlLCAiYWdlbnRfcm9sZSI6ICJGcmll
+      bmRseSBOZWlnaGJvciIsICJhZ2VudF9rZXkiOiAiOThmM2IxZDQ3Y2U5NjljZjA1NzcyN2I3ODQx
+      NDI1Y2QiLCAidG9vbHNfbmFtZXMiOiBbImRlY2lkZSBncmVldGluZ3MiXX1degIYAYUBAAEAABKO
+      AgoQnjTp5boK7/+DQxztYIpqihIIgGnMUkBtzHEqDFRhc2sgQ3JlYXRlZDABOcpYcuRvKBUYQalE
+      c+RvKBUYSi4KCGNyZXdfa2V5EiIKIGMzMDc2MDA5MzI2NzYxNDQ0ZDU3YzcxZDFkYTNmMjdjSjEK
+      B2NyZXdfaWQSJgokNzk0MDE5MjUtYjhlOS00NzA4LTg1MzAtNDQ4YWZhM2JmOGIwSi4KCHRhc2tf
+      a2V5EiIKIDgwZDdiY2Q0OTA5OTI5MDA4MzgzMmYwZTk4MzM4MGRmSjEKB3Rhc2tfaWQSJgokMmZm
+      NjE5N2UtYmEyNy00YjczLWI0YTctNGZhMDQ4ZTYyYjQ3egIYAYUBAAEAABKTAQoQ26H9pLUgswDN
+      p9XhJwwL6BIIx3bw7mAvPYwqClRvb2wgVXNhZ2UwATmy7NPlbygVGEEvb+HlbygVGEoaCg5jcmV3
+      YWlfdmVyc2lvbhIICgYwLjg2LjBKHwoJdG9vbF9uYW1lEhIKEERlY2lkZSBHcmVldGluZ3NKDgoI
+      YXR0ZW1wdHMSAhgBegIYAYUBAAEAAA==
+    headers:
+      Accept:
+      - '*/*'
+      Accept-Encoding:
+      - gzip, deflate
+      Connection:
+      - keep-alive
+      Content-Length:
+      - '2986'
+      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:
+      - Fri, 27 Dec 2024 22:14:53 GMT
+    status:
+      code: 200
+      message: OK
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are test role. test backstory\nYour
+      personal goal is: test goal\nTo give my best complete final answer to the task
+      use 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: Say the word: Hi\n\nThis
+      is the expect criteria for your final answer: The word: 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:"],
+      "stream": false}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '824'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=ePJSDFdHag2D8lj21_ijAMWjoA6xfnPNxN4uekvC728-1727226247743-0.0.1.1-604800000
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.52.1
+      x-stainless-arch:
+      - x64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - Linux
+      x-stainless-package-version:
+      - 1.52.1
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.7
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AjCtZLLrWi8ZASpP9bz6HaCV7xBIn\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1735337693,\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\":
+      158,\n    \"completion_tokens\": 12,\n    \"total_tokens\": 170,\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  \"system_fingerprint\":
+      \"fp_0aa8d3e20b\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 8f8caa83deca756b-SEA
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Fri, 27 Dec 2024 22:14:53 GMT
+      Server:
+      - cloudflare
+      Set-Cookie:
+      - __cf_bm=wJkq_yLkzE3OdxE0aMJz.G0kce969.9JxRmZ0ratl4c-1735337693-1.0.1.1-OKpUoRrSPFGvWv5Hp5ET1PNZ7iZNHPKEAuakpcQUxxPSeisUIIR3qIOZ31MGmYugqB5.wkvidgbxOAagqJvmnw;
+        path=/; expires=Fri, 27-Dec-24 22:44:53 GMT; domain=.api.openai.com; HttpOnly;
+        Secure; SameSite=None
+      - _cfuvid=A_ASCLNAVfQoyucWOAIhecWtEpNotYoZr0bAFihgNxs-1735337693273-0.0.1.1-604800000;
+        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:
+      - '404'
+      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:
+      - '149999816'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_6ac84634bff9193743c4b0911c09b4a6
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "system", "content": "Determine if the following
+      feedback indicates that the user is satisfied or if further changes are needed.
+      Respond with ''True'' if further changes are needed, or ''False'' if the user
+      is satisfied. **Important** Do not include any additional commentary outside
+      of your ''True'' or ''False'' response.\n\nFeedback: \"Don''t say hi, say Hello
+      instead!\""}], "model": "gpt-4o-mini", "stop": ["\nObservation:"], "stream":
+      false}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '461'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=A_ASCLNAVfQoyucWOAIhecWtEpNotYoZr0bAFihgNxs-1735337693273-0.0.1.1-604800000;
+        __cf_bm=wJkq_yLkzE3OdxE0aMJz.G0kce969.9JxRmZ0ratl4c-1735337693-1.0.1.1-OKpUoRrSPFGvWv5Hp5ET1PNZ7iZNHPKEAuakpcQUxxPSeisUIIR3qIOZ31MGmYugqB5.wkvidgbxOAagqJvmnw
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.52.1
+      x-stainless-arch:
+      - x64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - Linux
+      x-stainless-package-version:
+      - 1.52.1
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.7
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AjCtZNlWdrrPZhq0MJDqd16sMuQEJ\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1735337693,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"True\",\n        \"refusal\": null\n
+      \     },\n      \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n
+      \ ],\n  \"usage\": {\n    \"prompt_tokens\": 78,\n    \"completion_tokens\":
+      1,\n    \"total_tokens\": 79,\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  \"system_fingerprint\":
+      \"fp_0aa8d3e20b\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 8f8caa87094f756b-SEA
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Fri, 27 Dec 2024 22:14:53 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:
+      - '156'
+      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:
+      - '149999898'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_ec74bef2a9ef7b2144c03fd7f7bbeab0
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "system", "content": "You are test role. test backstory\nYour
+      personal goal is: test goal\nTo give my best complete final answer to the task
+      use 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: Say the word: Hi\n\nThis
+      is the expect criteria for your final answer: The word: 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": "I now
+      can give a great answer  \nFinal Answer: Hi"}, {"role": "user", "content": "Feedback:
+      Don''t say hi, say Hello instead!"}], "model": "gpt-4o-mini", "stop": ["\nObservation:"],
+      "stream": false}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '986'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=A_ASCLNAVfQoyucWOAIhecWtEpNotYoZr0bAFihgNxs-1735337693273-0.0.1.1-604800000;
+        __cf_bm=wJkq_yLkzE3OdxE0aMJz.G0kce969.9JxRmZ0ratl4c-1735337693-1.0.1.1-OKpUoRrSPFGvWv5Hp5ET1PNZ7iZNHPKEAuakpcQUxxPSeisUIIR3qIOZ31MGmYugqB5.wkvidgbxOAagqJvmnw
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.52.1
+      x-stainless-arch:
+      - x64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - Linux
+      x-stainless-package-version:
+      - 1.52.1
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.7
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AjCtZGv4f3h7GDdhyOy9G0sB1lRgC\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1735337693,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"Thought: I understand the feedback and
+      will adjust my response accordingly.  \\nFinal Answer: Hello\",\n        \"refusal\":
+      null\n      },\n      \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n
+      \   }\n  ],\n  \"usage\": {\n    \"prompt_tokens\": 188,\n    \"completion_tokens\":
+      18,\n    \"total_tokens\": 206,\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  \"system_fingerprint\":
+      \"fp_0aa8d3e20b\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 8f8caa88cac4756b-SEA
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Fri, 27 Dec 2024 22:14:54 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:
+      - '358'
+      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:
+      - '149999793'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_ae1ab6b206d28ded6fee3c83ed0c2ab7
+    http_version: HTTP/1.1
+    status_code: 200
+- request:
+    body: '{"messages": [{"role": "system", "content": "Determine if the following
+      feedback indicates that the user is satisfied or if further changes are needed.
+      Respond with ''True'' if further changes are needed, or ''False'' if the user
+      is satisfied. **Important** Do not include any additional commentary outside
+      of your ''True'' or ''False'' response.\n\nFeedback: \"looks good\""}], "model":
+      "gpt-4o-mini", "stop": ["\nObservation:"], "stream": false}'
+    headers:
+      accept:
+      - application/json
+      accept-encoding:
+      - gzip, deflate
+      connection:
+      - keep-alive
+      content-length:
+      - '439'
+      content-type:
+      - application/json
+      cookie:
+      - _cfuvid=A_ASCLNAVfQoyucWOAIhecWtEpNotYoZr0bAFihgNxs-1735337693273-0.0.1.1-604800000;
+        __cf_bm=wJkq_yLkzE3OdxE0aMJz.G0kce969.9JxRmZ0ratl4c-1735337693-1.0.1.1-OKpUoRrSPFGvWv5Hp5ET1PNZ7iZNHPKEAuakpcQUxxPSeisUIIR3qIOZ31MGmYugqB5.wkvidgbxOAagqJvmnw
+      host:
+      - api.openai.com
+      user-agent:
+      - OpenAI/Python 1.52.1
+      x-stainless-arch:
+      - x64
+      x-stainless-async:
+      - 'false'
+      x-stainless-lang:
+      - python
+      x-stainless-os:
+      - Linux
+      x-stainless-package-version:
+      - 1.52.1
+      x-stainless-raw-response:
+      - 'true'
+      x-stainless-retry-count:
+      - '0'
+      x-stainless-runtime:
+      - CPython
+      x-stainless-runtime-version:
+      - 3.12.7
+    method: POST
+    uri: https://api.openai.com/v1/chat/completions
+  response:
+    content: "{\n  \"id\": \"chatcmpl-AjCtaiHL4TY8Dssk0j2miqmjrzquy\",\n  \"object\":
+      \"chat.completion\",\n  \"created\": 1735337694,\n  \"model\": \"gpt-4o-mini-2024-07-18\",\n
+      \ \"choices\": [\n    {\n      \"index\": 0,\n      \"message\": {\n        \"role\":
+      \"assistant\",\n        \"content\": \"False\",\n        \"refusal\": null\n
+      \     },\n      \"logprobs\": null,\n      \"finish_reason\": \"stop\"\n    }\n
+      \ ],\n  \"usage\": {\n    \"prompt_tokens\": 73,\n    \"completion_tokens\":
+      1,\n    \"total_tokens\": 74,\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  \"system_fingerprint\":
+      \"fp_0aa8d3e20b\"\n}\n"
+    headers:
+      CF-Cache-Status:
+      - DYNAMIC
+      CF-RAY:
+      - 8f8caa8bdd26756b-SEA
+      Connection:
+      - keep-alive
+      Content-Encoding:
+      - gzip
+      Content-Type:
+      - application/json
+      Date:
+      - Fri, 27 Dec 2024 22:14:54 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:
+      - '184'
+      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:
+      - '149999902'
+      x-ratelimit-reset-requests:
+      - 2ms
+      x-ratelimit-reset-tokens:
+      - 0s
+      x-request-id:
+      - req_652891f79c1104a7a8436275d78a69f1
+    http_version: HTTP/1.1
+    status_code: 200
+version: 1

tests/test_task_guardrails.py

@@ -1,129 +1,179 @@
 """Tests for task guardrails functionality."""
 
+from typing import Any, Dict
 from unittest.mock import Mock
 
 import pytest
 
 from crewai.task import Task
+from crewai.tasks.exceptions import GuardrailValidationError
 from crewai.tasks.task_output import TaskOutput
 
 
-def test_task_without_guardrail():
-    """Test that tasks work normally without guardrails (backward compatibility)."""
-    agent = Mock()
-    agent.role = "test_agent"
-    agent.execute_task.return_value = "test result"
-    agent.crew = None
+class TestTaskGuardrails:
+    """Test suite for task guardrail functionality."""
 
-    task = Task(description="Test task", expected_output="Output")
+    @pytest.fixture
+    def mock_agent(self):
+        """Fixture providing a mock agent for testing."""
+        agent = Mock()
+        agent.role = "test_agent"
+        agent.crew = None
+        return agent
 
-    result = task.execute_sync(agent=agent)
-    assert isinstance(result, TaskOutput)
-    assert result.raw == "test result"
+    def test_task_without_guardrail(self, mock_agent):
+        """Test that tasks work normally without guardrails (backward compatibility)."""
+        mock_agent.execute_task.return_value = "test result"
+        task = Task(description="Test task", expected_output="Output")
+
+        result = task.execute_sync(agent=mock_agent)
+        assert isinstance(result, TaskOutput)
+        assert result.raw == "test result"
 
 
-def test_task_with_successful_guardrail():
-    """Test that successful guardrail validation passes transformed result."""
+    def test_task_with_successful_guardrail(self, mock_agent):
+        """Test that successful guardrail validation passes transformed result."""
+        def guardrail(result: TaskOutput):
+            return (True, result.raw.upper())
 
-    def guardrail(result: TaskOutput):
-        return (True, result.raw.upper())
+        mock_agent.execute_task.return_value = "test result"
+        task = Task(description="Test task", expected_output="Output", guardrail=guardrail)
 
-    agent = Mock()
-    agent.role = "test_agent"
-    agent.execute_task.return_value = "test result"
-    agent.crew = None
-
-    task = Task(description="Test task", expected_output="Output", guardrail=guardrail)
-
-    result = task.execute_sync(agent=agent)
-    assert isinstance(result, TaskOutput)
-    assert result.raw == "TEST RESULT"
+        result = task.execute_sync(agent=mock_agent)
+        assert isinstance(result, TaskOutput)
+        assert result.raw == "TEST RESULT"
 
 
-def test_task_with_failing_guardrail():
-    """Test that failing guardrail triggers retry with error context."""
+    def test_task_with_failing_guardrail(self, mock_agent):
+        """Test that failing guardrail triggers retry with error context."""
+        def guardrail(result: TaskOutput):
+            return (False, "Invalid format")
 
-    def guardrail(result: TaskOutput):
-        return (False, "Invalid format")
+        mock_agent.execute_task.side_effect = ["bad result", "good result"]
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail,
+            max_retries=1,
+        )
 
-    agent = Mock()
-    agent.role = "test_agent"
-    agent.execute_task.side_effect = ["bad result", "good result"]
-    agent.crew = None
+        # First execution fails guardrail, second succeeds
+        mock_agent.execute_task.side_effect = ["bad result", "good result"]
+        with pytest.raises(Exception) as exc_info:
+            task.execute_sync(agent=mock_agent)
 
-    task = Task(
-        description="Test task",
-        expected_output="Output",
-        guardrail=guardrail,
-        max_retries=1,
-    )
-
-    # First execution fails guardrail, second succeeds
-    agent.execute_task.side_effect = ["bad result", "good result"]
-    with pytest.raises(Exception) as exc_info:
-        task.execute_sync(agent=agent)
-
-    assert "Task failed guardrail validation" in str(exc_info.value)
-    assert task.retry_count == 1
+        assert "Task failed guardrail validation" in str(exc_info.value)
+        assert task.retry_count == 1
 
 
-def test_task_with_guardrail_retries():
-    """Test that guardrail respects max_retries configuration."""
+    def test_task_with_guardrail_retries(self, mock_agent):
+        """Test that guardrail respects max_retries configuration."""
+        def guardrail(result: TaskOutput):
+            return (False, "Invalid format")
 
-    def guardrail(result: TaskOutput):
-        return (False, "Invalid format")
+        mock_agent.execute_task.return_value = "bad result"
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail,
+            max_retries=2,
+        )
 
-    agent = Mock()
-    agent.role = "test_agent"
-    agent.execute_task.return_value = "bad result"
-    agent.crew = None
+        with pytest.raises(Exception) as exc_info:
+            task.execute_sync(agent=mock_agent)
 
-    task = Task(
-        description="Test task",
-        expected_output="Output",
-        guardrail=guardrail,
-        max_retries=2,
-    )
-
-    with pytest.raises(Exception) as exc_info:
-        task.execute_sync(agent=agent)
-
-    assert task.retry_count == 2
-    assert "Task failed guardrail validation after 2 retries" in str(exc_info.value)
-    assert "Invalid format" in str(exc_info.value)
+        assert task.retry_count == 2
+        assert "Task failed guardrail validation after 2 retries" in str(exc_info.value)
+        assert "Invalid format" in str(exc_info.value)
 
 
-def test_guardrail_error_in_context():
-    """Test that guardrail error is passed in context for retry."""
+    def test_guardrail_error_in_context(self, mock_agent):
+        """Test that guardrail error is passed in context for retry."""
+        def guardrail(result: TaskOutput):
+            return (False, "Expected JSON, got string")
 
-    def guardrail(result: TaskOutput):
-        return (False, "Expected JSON, got string")
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail,
+            max_retries=1,
+        )
 
-    agent = Mock()
-    agent.role = "test_agent"
-    agent.crew = None
+        # Mock execute_task to succeed on second attempt
+        first_call = True
+        def execute_task(task, context, tools):
+            nonlocal first_call
+            if first_call:
+                first_call = False
+                return "invalid"
+            return '{"valid": "json"}'
 
-    task = Task(
-        description="Test task",
-        expected_output="Output",
-        guardrail=guardrail,
-        max_retries=1,
-    )
+        mock_agent.execute_task.side_effect = execute_task
 
-    # Mock execute_task to succeed on second attempt
-    first_call = True
+        with pytest.raises(Exception) as exc_info:
+            task.execute_sync(agent=mock_agent)
 
-    def execute_task(task, context, tools):
-        nonlocal first_call
-        if first_call:
-            first_call = False
-            return "invalid"
-        return '{"valid": "json"}'
+        assert "Task failed guardrail validation" in str(exc_info.value)
+        assert "Expected JSON, got string" in str(exc_info.value)
 
-    agent.execute_task.side_effect = execute_task
 
-    with pytest.raises(Exception) as exc_info:
-        task.execute_sync(agent=agent)
+    def test_guardrail_with_new_style_annotation(self, mock_agent):
+        """Test guardrail with new style tuple annotation."""
+        def guardrail(result: TaskOutput) -> tuple[bool, str]:
+            return (True, result.raw.upper())
+        
+        mock_agent.execute_task.return_value = "test result"
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail
+        )
 
-    assert "Task failed guardrail validation" in str(exc_info.value)
-    assert "Expected JSON, got string" in str(exc_info.value)
+        result = task.execute_sync(agent=mock_agent)
+        assert isinstance(result, TaskOutput)
+        assert result.raw == "TEST RESULT"
+
+    def test_guardrail_with_optional_params(self, mock_agent):
+        """Test guardrail with optional parameters."""
+        def guardrail(result: TaskOutput, optional_param: str = "default") -> tuple[bool, str]:
+            return (True, f"{result.raw}-{optional_param}")
+        
+        mock_agent.execute_task.return_value = "test"
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail
+        )
+
+        result = task.execute_sync(agent=mock_agent)
+        assert isinstance(result, TaskOutput)
+        assert result.raw == "test-default"
+
+    def test_guardrail_with_invalid_optional_params(self, mock_agent):
+        """Test guardrail with invalid optional parameters."""
+        def guardrail(result: TaskOutput, *, required_kwonly: str) -> tuple[bool, str]:
+            return (True, result.raw)
+        
+        with pytest.raises(GuardrailValidationError) as exc_info:
+            Task(
+                description="Test task",
+                expected_output="Output",
+                guardrail=guardrail
+            )
+        assert "exactly one required positional parameter" in str(exc_info.value)
+
+    def test_guardrail_with_dict_return_type(self, mock_agent):
+        """Test guardrail with dict return type."""
+        def guardrail(result: TaskOutput) -> tuple[bool, dict[str, Any]]:
+            return (True, {"processed": result.raw.upper()})
+        
+        mock_agent.execute_task.return_value = "test"
+        task = Task(
+            description="Test task",
+            expected_output="Output",
+            guardrail=guardrail
+        )
+
+        result = task.execute_sync(agent=mock_agent)
+        assert isinstance(result, TaskOutput)
+        assert result.raw == {"processed": "TEST"}

tests/utilities/cassettes/test_converter_with_llama3_2_model.yaml

@@ -1,9 +1,14 @@
 interactions:
 - request:
-    body: '{"model": "llama3.2:3b", "prompt": "### System:\nPlease convert the following
-      text into valid JSON.\n\nOutput ONLY the valid JSON and nothing else.\n\nThe
-      JSON must follow this format exactly:\n{\n  \"name\": str,\n  \"age\": int\n}\n\n###
-      User:\nName: Alice Llama, Age: 30\n\n", "options": {"stop": []}, "stream": false}'
+    body: '{"model": "llama3.2:3b", "prompt": "### User:\nName: Alice Llama, Age:
+      30\n\n### System:\nProduce JSON OUTPUT ONLY! Adhere to this format {\"name\":
+      \"function_name\", \"arguments\":{\"argument_name\": \"argument_value\"}} The
+      following functions are available to you:\n{''type'': ''function'', ''function'':
+      {''name'': ''SimpleModel'', ''description'': ''Correctly extracted `SimpleModel`
+      with all the required parameters with correct types'', ''parameters'': {''properties'':
+      {''name'': {''title'': ''Name'', ''type'': ''string''}, ''age'': {''title'':
+      ''Age'', ''type'': ''integer''}}, ''required'': [''age'', ''name''], ''type'':
+      ''object''}}}\n\n\n", "options": {}, "stream": false, "format": "json"}'
     headers:
       accept:
       - '*/*'
@@ -12,23 +17,23 @@ interactions:
       connection:
       - keep-alive
       content-length:
-      - '321'
+      - '657'
       host:
       - localhost:11434
       user-agent:
-      - litellm/1.60.2
+      - litellm/1.57.4
     method: POST
     uri: http://localhost:11434/api/generate
   response:
-    content: '{"model":"llama3.2:3b","created_at":"2025-02-21T02:57:55.059392Z","response":"{\"name\":
-      \"Alice Llama\", \"age\": 30}","done":true,"done_reason":"stop","context":[128006,9125,128007,271,38766,1303,33025,2696,25,6790,220,2366,18,271,128009,128006,882,128007,271,14711,744,512,5618,5625,279,2768,1495,1139,2764,4823,382,5207,27785,279,2764,4823,323,4400,775,382,791,4823,2011,1833,420,3645,7041,512,517,220,330,609,794,610,345,220,330,425,794,528,198,633,14711,2724,512,678,25,30505,445,81101,11,13381,25,220,966,271,128009,128006,78191,128007,271,5018,609,794,330,62786,445,81101,498,330,425,794,220,966,92],"total_duration":4675906000,"load_duration":836091458,"prompt_eval_count":82,"prompt_eval_duration":3561000000,"eval_count":15,"eval_duration":275000000}'
+    content: '{"model":"llama3.2:3b","created_at":"2025-01-15T20:47:11.926411Z","response":"{\"name\":
+      \"SimpleModel\", \"arguments\":{\"name\": \"Alice Llama\", \"age\": 30}}","done":true,"done_reason":"stop","context":[128006,9125,128007,271,38766,1303,33025,2696,25,6790,220,2366,18,271,128009,128006,882,128007,271,14711,2724,512,678,25,30505,445,81101,11,13381,25,220,966,271,14711,744,512,1360,13677,4823,32090,27785,0,2467,6881,311,420,3645,5324,609,794,330,1723,1292,498,330,16774,23118,14819,1292,794,330,14819,3220,32075,578,2768,5865,527,2561,311,499,512,13922,1337,1232,364,1723,518,364,1723,1232,5473,609,1232,364,16778,1747,518,364,4789,1232,364,34192,398,28532,1595,16778,1747,63,449,682,279,2631,5137,449,4495,4595,518,364,14105,1232,5473,13495,1232,5473,609,1232,5473,2150,1232,364,678,518,364,1337,1232,364,928,25762,364,425,1232,5473,2150,1232,364,17166,518,364,1337,1232,364,11924,8439,2186,364,6413,1232,2570,425,518,364,609,4181,364,1337,1232,364,1735,23742,3818,128009,128006,78191,128007,271,5018,609,794,330,16778,1747,498,330,16774,23118,609,794,330,62786,445,81101,498,330,425,794,220,966,3500],"total_duration":3374470708,"load_duration":1075750500,"prompt_eval_count":167,"prompt_eval_duration":1871000000,"eval_count":24,"eval_duration":426000000}'
     headers:
       Content-Length:
-      - '761'
+      - '1263'
       Content-Type:
       - application/json; charset=utf-8
       Date:
-      - Fri, 21 Feb 2025 02:57:55 GMT
+      - Wed, 15 Jan 2025 20:47:12 GMT
     http_version: HTTP/1.1
     status_code: 200
 - request:
@@ -47,7 +52,7 @@ interactions:
       host:
       - localhost:11434
       user-agent:
-      - litellm/1.60.2
+      - litellm/1.57.4
     method: POST
     uri: http://localhost:11434/api/show
   response:
@@ -223,7 +228,7 @@ interactions:
       Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama
       3.2: LlamaUseReport@meta.com\",\"modelfile\":\"# Modelfile generated by \\\"ollama
       show\\\"\\n# To build a new Modelfile based on this, replace FROM with:\\n#
-      FROM llama3.2:3b\\n\\nFROM /Users/joaomoura/.ollama/models/blobs/sha256-dde5aa3fc5ffc17176b5e8bdc82f587b24b2678c6c66101bf7da77af9f7ccdff\\nTEMPLATE
+      FROM llama3.2:3b\\n\\nFROM /Users/brandonhancock/.ollama/models/blobs/sha256-dde5aa3fc5ffc17176b5e8bdc82f587b24b2678c6c66101bf7da77af9f7ccdff\\nTEMPLATE
       \\\"\\\"\\\"\\u003c|start_header_id|\\u003esystem\\u003c|end_header_id|\\u003e\\n\\nCutting
       Knowledge Date: December 2023\\n\\n{{ if .System }}{{ .System }}\\n{{- end }}\\n{{-
       if .Tools }}When you receive a tool call response, use the output to format
@@ -436,12 +441,12 @@ interactions:
       .Content }}\\n{{- end }}{{ if not $last }}\\u003c|eot_id|\\u003e{{ end }}\\n{{-
       else if eq .Role \\\"tool\\\" }}\\u003c|start_header_id|\\u003eipython\\u003c|end_header_id|\\u003e\\n\\n{{
       .Content }}\\u003c|eot_id|\\u003e{{ if $last }}\\u003c|start_header_id|\\u003eassistant\\u003c|end_header_id|\\u003e\\n\\n{{
-      end }}\\n{{- end }}\\n{{- end }}\",\"details\":{\"parent_model\":\"\",\"format\":\"gguf\",\"family\":\"llama\",\"families\":[\"llama\"],\"parameter_size\":\"3.2B\",\"quantization_level\":\"Q4_K_M\"},\"model_info\":{\"general.architecture\":\"llama\",\"general.basename\":\"Llama-3.2\",\"general.file_type\":15,\"general.finetune\":\"Instruct\",\"general.languages\":[\"en\",\"de\",\"fr\",\"it\",\"pt\",\"hi\",\"es\",\"th\"],\"general.parameter_count\":3212749888,\"general.quantization_version\":2,\"general.size_label\":\"3B\",\"general.tags\":[\"facebook\",\"meta\",\"pytorch\",\"llama\",\"llama-3\",\"text-generation\"],\"general.type\":\"model\",\"llama.attention.head_count\":24,\"llama.attention.head_count_kv\":8,\"llama.attention.key_length\":128,\"llama.attention.layer_norm_rms_epsilon\":0.00001,\"llama.attention.value_length\":128,\"llama.block_count\":28,\"llama.context_length\":131072,\"llama.embedding_length\":3072,\"llama.feed_forward_length\":8192,\"llama.rope.dimension_count\":128,\"llama.rope.freq_base\":500000,\"llama.vocab_size\":128256,\"tokenizer.ggml.bos_token_id\":128000,\"tokenizer.ggml.eos_token_id\":128009,\"tokenizer.ggml.merges\":null,\"tokenizer.ggml.model\":\"gpt2\",\"tokenizer.ggml.pre\":\"llama-bpe\",\"tokenizer.ggml.token_type\":null,\"tokenizer.ggml.tokens\":null},\"modified_at\":\"2025-02-20T18:55:09.150577031-08:00\"}"
+      end }}\\n{{- end }}\\n{{- end }}\",\"details\":{\"parent_model\":\"\",\"format\":\"gguf\",\"family\":\"llama\",\"families\":[\"llama\"],\"parameter_size\":\"3.2B\",\"quantization_level\":\"Q4_K_M\"},\"model_info\":{\"general.architecture\":\"llama\",\"general.basename\":\"Llama-3.2\",\"general.file_type\":15,\"general.finetune\":\"Instruct\",\"general.languages\":[\"en\",\"de\",\"fr\",\"it\",\"pt\",\"hi\",\"es\",\"th\"],\"general.parameter_count\":3212749888,\"general.quantization_version\":2,\"general.size_label\":\"3B\",\"general.tags\":[\"facebook\",\"meta\",\"pytorch\",\"llama\",\"llama-3\",\"text-generation\"],\"general.type\":\"model\",\"llama.attention.head_count\":24,\"llama.attention.head_count_kv\":8,\"llama.attention.key_length\":128,\"llama.attention.layer_norm_rms_epsilon\":0.00001,\"llama.attention.value_length\":128,\"llama.block_count\":28,\"llama.context_length\":131072,\"llama.embedding_length\":3072,\"llama.feed_forward_length\":8192,\"llama.rope.dimension_count\":128,\"llama.rope.freq_base\":500000,\"llama.vocab_size\":128256,\"tokenizer.ggml.bos_token_id\":128000,\"tokenizer.ggml.eos_token_id\":128009,\"tokenizer.ggml.merges\":null,\"tokenizer.ggml.model\":\"gpt2\",\"tokenizer.ggml.pre\":\"llama-bpe\",\"tokenizer.ggml.token_type\":null,\"tokenizer.ggml.tokens\":null},\"modified_at\":\"2024-12-31T11:53:14.529771974-05:00\"}"
     headers:
       Content-Type:
       - application/json; charset=utf-8
       Date:
-      - Fri, 21 Feb 2025 02:57:55 GMT
+      - Wed, 15 Jan 2025 20:47:12 GMT
       Transfer-Encoding:
       - chunked
     http_version: HTTP/1.1
@@ -462,7 +467,7 @@ interactions:
       host:
       - localhost:11434
       user-agent:
-      - litellm/1.60.2
+      - litellm/1.57.4
     method: POST
     uri: http://localhost:11434/api/show
   response:
@@ -638,7 +643,7 @@ interactions:
       Reporting violations of the Acceptable Use Policy or unlicensed uses of Llama
       3.2: LlamaUseReport@meta.com\",\"modelfile\":\"# Modelfile generated by \\\"ollama
       show\\\"\\n# To build a new Modelfile based on this, replace FROM with:\\n#
-      FROM llama3.2:3b\\n\\nFROM /Users/joaomoura/.ollama/models/blobs/sha256-dde5aa3fc5ffc17176b5e8bdc82f587b24b2678c6c66101bf7da77af9f7ccdff\\nTEMPLATE
+      FROM llama3.2:3b\\n\\nFROM /Users/brandonhancock/.ollama/models/blobs/sha256-dde5aa3fc5ffc17176b5e8bdc82f587b24b2678c6c66101bf7da77af9f7ccdff\\nTEMPLATE
       \\\"\\\"\\\"\\u003c|start_header_id|\\u003esystem\\u003c|end_header_id|\\u003e\\n\\nCutting
       Knowledge Date: December 2023\\n\\n{{ if .System }}{{ .System }}\\n{{- end }}\\n{{-
       if .Tools }}When you receive a tool call response, use the output to format
@@ -851,12 +856,12 @@ interactions:
       .Content }}\\n{{- end }}{{ if not $last }}\\u003c|eot_id|\\u003e{{ end }}\\n{{-
       else if eq .Role \\\"tool\\\" }}\\u003c|start_header_id|\\u003eipython\\u003c|end_header_id|\\u003e\\n\\n{{
       .Content }}\\u003c|eot_id|\\u003e{{ if $last }}\\u003c|start_header_id|\\u003eassistant\\u003c|end_header_id|\\u003e\\n\\n{{
-      end }}\\n{{- end }}\\n{{- end }}\",\"details\":{\"parent_model\":\"\",\"format\":\"gguf\",\"family\":\"llama\",\"families\":[\"llama\"],\"parameter_size\":\"3.2B\",\"quantization_level\":\"Q4_K_M\"},\"model_info\":{\"general.architecture\":\"llama\",\"general.basename\":\"Llama-3.2\",\"general.file_type\":15,\"general.finetune\":\"Instruct\",\"general.languages\":[\"en\",\"de\",\"fr\",\"it\",\"pt\",\"hi\",\"es\",\"th\"],\"general.parameter_count\":3212749888,\"general.quantization_version\":2,\"general.size_label\":\"3B\",\"general.tags\":[\"facebook\",\"meta\",\"pytorch\",\"llama\",\"llama-3\",\"text-generation\"],\"general.type\":\"model\",\"llama.attention.head_count\":24,\"llama.attention.head_count_kv\":8,\"llama.attention.key_length\":128,\"llama.attention.layer_norm_rms_epsilon\":0.00001,\"llama.attention.value_length\":128,\"llama.block_count\":28,\"llama.context_length\":131072,\"llama.embedding_length\":3072,\"llama.feed_forward_length\":8192,\"llama.rope.dimension_count\":128,\"llama.rope.freq_base\":500000,\"llama.vocab_size\":128256,\"tokenizer.ggml.bos_token_id\":128000,\"tokenizer.ggml.eos_token_id\":128009,\"tokenizer.ggml.merges\":null,\"tokenizer.ggml.model\":\"gpt2\",\"tokenizer.ggml.pre\":\"llama-bpe\",\"tokenizer.ggml.token_type\":null,\"tokenizer.ggml.tokens\":null},\"modified_at\":\"2025-02-20T18:55:09.150577031-08:00\"}"
+      end }}\\n{{- end }}\\n{{- end }}\",\"details\":{\"parent_model\":\"\",\"format\":\"gguf\",\"family\":\"llama\",\"families\":[\"llama\"],\"parameter_size\":\"3.2B\",\"quantization_level\":\"Q4_K_M\"},\"model_info\":{\"general.architecture\":\"llama\",\"general.basename\":\"Llama-3.2\",\"general.file_type\":15,\"general.finetune\":\"Instruct\",\"general.languages\":[\"en\",\"de\",\"fr\",\"it\",\"pt\",\"hi\",\"es\",\"th\"],\"general.parameter_count\":3212749888,\"general.quantization_version\":2,\"general.size_label\":\"3B\",\"general.tags\":[\"facebook\",\"meta\",\"pytorch\",\"llama\",\"llama-3\",\"text-generation\"],\"general.type\":\"model\",\"llama.attention.head_count\":24,\"llama.attention.head_count_kv\":8,\"llama.attention.key_length\":128,\"llama.attention.layer_norm_rms_epsilon\":0.00001,\"llama.attention.value_length\":128,\"llama.block_count\":28,\"llama.context_length\":131072,\"llama.embedding_length\":3072,\"llama.feed_forward_length\":8192,\"llama.rope.dimension_count\":128,\"llama.rope.freq_base\":500000,\"llama.vocab_size\":128256,\"tokenizer.ggml.bos_token_id\":128000,\"tokenizer.ggml.eos_token_id\":128009,\"tokenizer.ggml.merges\":null,\"tokenizer.ggml.model\":\"gpt2\",\"tokenizer.ggml.pre\":\"llama-bpe\",\"tokenizer.ggml.token_type\":null,\"tokenizer.ggml.tokens\":null},\"modified_at\":\"2024-12-31T11:53:14.529771974-05:00\"}"
     headers:
       Content-Type:
       - application/json; charset=utf-8
       Date:
-      - Fri, 21 Feb 2025 02:57:55 GMT
+      - Wed, 15 Jan 2025 20:47:12 GMT
       Transfer-Encoding:
       - chunked
     http_version: HTTP/1.1

tests/utilities/test_converter.py

@@ -1,5 +1,4 @@
 import json
-import os
 from typing import Dict, List, Optional
 from unittest.mock import MagicMock, Mock, patch
 
@@ -221,13 +220,10 @@ def test_get_conversion_instructions_gpt():
         supports_function_calling.return_value = True
         instructions = get_conversion_instructions(SimpleModel, llm)
         model_schema = PydanticSchemaParser(model=SimpleModel).get_schema()
-        expected_instructions = (
-            "Please convert the following text into valid JSON.\n\n"
-            "Output ONLY the valid JSON and nothing else.\n\n"
-            "The JSON must follow this schema exactly:\n```json\n"
-            f"{model_schema}\n```"
+        assert (
+            instructions
+            == f"Please convert the following text into valid JSON.\n\nThe JSON should follow this schema:\n```json\n{model_schema}\n```"
         )
-        assert instructions == expected_instructions
 
 
 def test_get_conversion_instructions_non_gpt():
@@ -350,17 +346,12 @@ def test_convert_with_instructions():
     assert output.age == 30
 
 
-# Skip tests that call external APIs when running in CI/CD
-skip_external_api = pytest.mark.skipif(
-    os.getenv("CI") is not None, reason="Skipping tests that call external API in CI/CD"
-)
-
-
-@skip_external_api
-@pytest.mark.vcr(filter_headers=["authorization"], record_mode="once")
+@pytest.mark.vcr(filter_headers=["authorization"])
 def test_converter_with_llama3_2_model():
     llm = LLM(model="ollama/llama3.2:3b", base_url="http://localhost:11434")
+
     sample_text = "Name: Alice Llama, Age: 30"
+
     instructions = get_conversion_instructions(SimpleModel, llm)
     converter = Converter(
         llm=llm,
@@ -368,17 +359,19 @@ def test_converter_with_llama3_2_model():
         model=SimpleModel,
         instructions=instructions,
     )
+
     output = converter.to_pydantic()
+
     assert isinstance(output, SimpleModel)
     assert output.name == "Alice Llama"
     assert output.age == 30
 
 
-@skip_external_api
-@pytest.mark.vcr(filter_headers=["authorization"], record_mode="once")
+@pytest.mark.vcr(filter_headers=["authorization"])
 def test_converter_with_llama3_1_model():
     llm = LLM(model="ollama/llama3.1", base_url="http://localhost:11434")
     sample_text = "Name: Alice Llama, Age: 30"
+
     instructions = get_conversion_instructions(SimpleModel, llm)
     converter = Converter(
         llm=llm,
@@ -386,19 +379,14 @@ def test_converter_with_llama3_1_model():
         model=SimpleModel,
         instructions=instructions,
     )
+
     output = converter.to_pydantic()
+
     assert isinstance(output, SimpleModel)
     assert output.name == "Alice Llama"
     assert output.age == 30
 
 
-# Skip tests that call external APIs when running in CI/CD
-skip_external_api = pytest.mark.skipif(
-    os.getenv("CI") is not None, reason="Skipping tests that call external API in CI/CD"
-)
-
-
-@skip_external_api
 @pytest.mark.vcr(filter_headers=["authorization"])
 def test_converter_with_nested_model():
     llm = LLM(model="gpt-4o-mini")
@@ -575,7 +563,7 @@ def test_converter_with_ambiguous_input():
     with pytest.raises(ConverterError) as exc_info:
         output = converter.to_pydantic()
 
-    assert "failed to convert text into a pydantic model" in str(exc_info.value).lower()
+    assert "validation error" in str(exc_info.value).lower()
 
 
 # Tests for function calling support

tests/utilities/test_token_process.py

@@ -1,49 +0,0 @@
-import unittest
-
-from crewai.agents.agent_builder.utilities.base_token_process import TokenProcess
-
-
-class TestTokenProcess(unittest.TestCase):
-    """Test suite for TokenProcess class token counting functionality."""
-
-    def setUp(self):
-        """Initialize a fresh TokenProcess instance before each test."""
-        self.token_process = TokenProcess()
-
-    def test_sum_cached_prompt_tokens_with_none(self):
-        """Test that passing None to sum_cached_prompt_tokens doesn't modify the counter."""
-        initial_tokens = self.token_process.cached_prompt_tokens
-        self.token_process.sum_cached_prompt_tokens(None)
-        self.assertEqual(self.token_process.cached_prompt_tokens, initial_tokens)
-
-    def test_sum_cached_prompt_tokens_with_int(self):
-        """Test that passing an integer correctly increments the counter."""
-        initial_tokens = self.token_process.cached_prompt_tokens
-        self.token_process.sum_cached_prompt_tokens(5)
-        self.assertEqual(self.token_process.cached_prompt_tokens, initial_tokens + 5)
-
-    def test_sum_cached_prompt_tokens_with_zero(self):
-        """Test that passing zero doesn't modify the counter."""
-        initial_tokens = self.token_process.cached_prompt_tokens
-        self.token_process.sum_cached_prompt_tokens(0)
-        self.assertEqual(self.token_process.cached_prompt_tokens, initial_tokens)
-
-    def test_sum_cached_prompt_tokens_with_large_number(self):
-        """Test that the counter works with large numbers."""
-        initial_tokens = self.token_process.cached_prompt_tokens
-        self.token_process.sum_cached_prompt_tokens(1000000)
-        self.assertEqual(self.token_process.cached_prompt_tokens, initial_tokens + 1000000)
-
-    def test_sum_cached_prompt_tokens_multiple_calls(self):
-        """Test that multiple calls accumulate correctly, ignoring None values."""
-        initial_tokens = self.token_process.cached_prompt_tokens
-        self.token_process.sum_cached_prompt_tokens(5)
-        self.token_process.sum_cached_prompt_tokens(None)
-        self.token_process.sum_cached_prompt_tokens(3)
-        self.assertEqual(self.token_process.cached_prompt_tokens, initial_tokens + 8)
-
-    def test_sum_cached_prompt_tokens_with_negative(self):
-        """Test that negative values raise ValueError."""
-        with self.assertRaises(ValueError) as context:
-            self.token_process.sum_cached_prompt_tokens(-1)
-        self.assertEqual(str(context.exception), "Token count cannot be negative")