Error while using a custom tool

[ravindra-bamboobox]

I am trying to use Crew AI agent with custom tool. Following is my code:

from crewai_tools import BaseTool

class reading_level_check(BaseTool):
    name: str = "reading_level_check"
    description: str = "This tool takes a text as input and returns it's reading ease as a float. If the reading ease number is 6, then the text can be read and understood by a 6th grade student."

    def _run(self, argument: str) -> str:
        return textstat.text_standard(argument, float_output=True)


os.environ["OPENAI_MODEL_NAME"] = 'gpt-4o-mini'

email_corrector = Agent(
role="Content Writer",
goal="Update / rewrite an email content so that it have reading level of 6th grade",
backstory="You are a content writer with 10 years of experience in a marketing team. "
          "Your task is to update / rewrite the provided email to make it simple to read. "
          "You should modify the email such that the reading level is that of 6th standard"
          "You must not generate false inforation or must not report information about which you are not sure.",
allow_delegation=False,
verbose=True
)

email_corrector_task = Task(
description=(
    "You will be provided a baseline email: {email_body}:\n"
    "You need to update / rewrite / modify the email to make it more simple to read. "
    "Your goal is to arrive at the reading level of 6th grade student "
    "Refer to following guidelines to increase the reading ease "
    "1. The text shall have smaller and simpler sentences. "
    "2. Each word within the sentence shall be simple / small. Avoid complicated words"
),
expected_output=(
    "Expected output is an updated email having reading ease of 6th grade student. "
    "You must not generate false inforation or must not report information about which you are not sure."
    "You should not make drastic changes to the email body unless they are necessary"
),
tools=[reading_level_check],
agent=email_corrector
)


crew = Crew(
        agents=[email_corrector],
        tasks=[email_corrector_task],
        verbose=True,
        memory=False
        )

result = crew.kickoff(inputs={"email_body": email})

I am getting the following error:

2024-08-30 15:51:41,924 - 11404 - __init__.py-__init__:531 - WARNING: Overriding of current TracerProvider is not allowed
 [2024-08-30 15:51:41][DEBUG]: == Working Agent: Content Writer
 [2024-08-30 15:51:41][INFO]: == Starting Task: You will be provided a baseline email: Subject: Struggling with Marketing ROI?

Hi Mary,

As someone who’s mastered the art of weaving through numbers and business growth at isolved, it’s inspiring to see how your expertise in integrating innovative technologies is shaping the future of HR management. Remembering your journey from Winthrop University to now, it’s clear that harnessing data for decision-making is your forte. Your focus on leveraging advanced analytics to better the employee experience is a testament to your commitment to excellence. At Bamboobox, we share your passion for data-driven insights and can’t help but appreciate the strides isolved has made in the realm of human capital management.

Speaking of data, have you considered how intent data might revolutionize your go-to-market strategies? Our latest insights delve into optimizing Account-Based Marketing by integrating such data. By aligning intent data with engagement metrics, you can craft personalized outreach efforts that speak directly to high-intent accounts. This approach not only enhances your team’s ability to prioritize but also shortens sales cycles and reveals new opportunities beyond the usual suspects. Discover how these strategies could potentially align with isolved's objectives by taking a closer look at our thoughts on breathing intent into your ABM program. 

Dive into the details here: https://bamboobox.ai/breathe-intent-into-your-abm-program/

Best regards,  
Susan  
BambooBox:
You need to update / rewrite / modify the email to make it more simple to read. Your goal is to arrive at the reading level of 6th grade student Refer to following guidelines to increase the reading ease 1. The text shall have smaller and simpler sentences. 2. Each word within the sentence shall be simple / small. Avoid complicated words
---------------------------------------------------------------------------
ValidationError                           Traceback (most recent call last)
Cell In[197], line 8
      1 crew = Crew(
      2         agents=[email_corrector],
      3         tasks=[email_corrector_task],
      4         verbose=True,
      5         memory=False
      6         )
----> 8 result = crew.kickoff(inputs={"email_body": email})

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\crew.py:469, in Crew.kickoff(self, inputs)
    466 metrics: List[UsageMetrics] = []
    468 if self.process == Process.sequential:
--> 469     result = self._run_sequential_process()
    470 elif self.process == Process.hierarchical:
    471     result = self._run_hierarchical_process()

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\crew.py:577, in Crew._run_sequential_process(self)
    575 def _run_sequential_process(self) -> CrewOutput:
    576     """Executes tasks sequentially and returns the final output."""
--> 577     return self._execute_tasks(self.tasks)

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\crew.py:665, in Crew._execute_tasks(self, tasks, start_index, was_replayed)
    662     futures.clear()
    664 context = self._get_context(task, task_outputs)
--> 665 task_output = task.execute_sync(
    666     agent=agent_to_use,
    667     context=context,
    668     tools=agent_to_use.tools,
    669 )
    670 task_outputs = [task_output]
    671 self._process_task_result(task, task_output)

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\task.py:180, in Task.execute_sync(self, agent, context, tools)
    173 def execute_sync(
    174     self,
    175     agent: Optional[BaseAgent] = None,
    176     context: Optional[str] = None,
    177     tools: Optional[List[Any]] = None,
    178 ) -> TaskOutput:
    179     """Execute the task synchronously."""
--> 180     return self._execute_core(agent, context, tools)

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\task.py:234, in Task._execute_core(self, agent, context, tools)
    231 self.prompt_context = context
    232 tools = tools or self.tools or []
--> 234 result = agent.execute_task(
    235     task=self,
    236     context=context,
    237     tools=tools,
    238 )
    240 pydantic_output, json_output = self._export_output(result)
    242 task_output = TaskOutput(
    243     description=self.description,
    244     raw=result,
   (...)
    248     output_format=self._get_output_format(),
    249 )

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\agent.py:188, in Agent.execute_task(self, task, context, tools)
    186 tools = tools or self.tools or []
    187 parsed_tools = self._parse_tools(tools)
--> 188 self.create_agent_executor(tools=tools)
    189 self.agent_executor.tools = parsed_tools
    190 self.agent_executor.task = task

File ~\AppData\Roaming\Python\Python311\site-packages\crewai\agent.py:305, in Agent.create_agent_executor(self, tools)
    302 bind = self.llm.bind(stop=stop_words)
    304 inner_agent = agent_args | execution_prompt | bind | CrewAgentParser(agent=self)
--> 305 self.agent_executor = CrewAgentExecutor(
    306     agent=RunnableAgent(runnable=inner_agent), **executor_args
    307 )

File ~\AppData\Roaming\Python\Python311\site-packages\langchain_core\load\serializable.py:113, in Serializable.__init__(self, *args, **kwargs)
    111 def __init__(self, *args: Any, **kwargs: Any) -> None:
    112     """"""
--> 113     super().__init__(*args, **kwargs)

File ~\AppData\Roaming\Python\Python311\site-packages\pydantic\v1\main.py:341, in BaseModel.__init__(__pydantic_self__, **data)
    339 values, fields_set, validation_error = validate_model(__pydantic_self__.__class__, data)
    340 if validation_error:
--> 341     raise validation_error
    342 try:
    343     object_setattr(__pydantic_self__, '__dict__', values)

ValidationError: 1 validation error for CrewAgentExecutor
tools -> 0
  value is not a valid dict (type=type_error.dict)

Here are the versions I am using:

Python: 3.11.5
crewAI: 0.51.1
crewAI tools: 0.8.3
langchain: 0.2.14
openai: 1.40.6
pydantic: 2.8.2

Anyone facing this issue? Would appreciate if anyone can help me out on this.