IMPORTANT ENV PACKAGE VERSIONS :-
Pydantic Version :- 2.9.2,
Crew Ai Version :- 0.76.2,
Python Version :- 3.10.15

Error : - Traceback (most recent call last): File "/Users/pravincoder/Developer/Email Advertiser/main.py", line 59, in <module> result = custom_crew.run() File "/Users/pravincoder/Developer/Email Advertiser/main.py", line 30, in run custom_task_1 = tasks.task_1_name( File "/Users/pravincoder/Developer/Email Advertiser/tasks.py", line 13, in task_1_name return Task( File "/opt/homebrew/Caskroom/miniconda/base/envs/email-crew/lib/python3.10/site-packages/pydantic/main.py", line 212, in __init__ validated_self = self.__pydantic_validator__.validate_python(data, self_instance=self) pydantic_core._pydantic_core.ValidationError: 1 validation error for Task expected_output Field required [type=missing, input_value={'description': "\nResear...agent 1 backstory here)}, input_type=dict] For further information visit https://errors.pydantic.dev/2.9/v/missing

Reason :- Task object requires an expected_output field, which you haven't provided. The Task model in pydantic expects specific fields to be filled based on how the Task class is defined.

Solution :- Add the expected_output field in your Task creation.
According Actual Solution:-

from crewai import Task
from textwrap import dedent


# This is an example of how to define custom tasks.
# You can define as many tasks as you want.
# You can also define custom agents in agents.py
class CustomTasks:
    def __tip_section(self):
        return "If you do your BEST WORK, I'll give you a $10,000 commission!"

    def task_1_name(self, agent, var1, var2):
        return Task(
            description=dedent(
                f"""
            Do something as part of task 1
            
            {self.__tip_section()}
    
            Make sure to use the most recent data as possible.
    
            Use this variable: {var1}
            And also this variable: {var2}
        """
            ),
        expected_output="Expected Output of this Agent(task_1_name) in simple words",
            agent=agent,
        )

    def task_2_name(self, agent):
        return Task(
            description=dedent(
                f"""
            Take the input from task 1 and do something with it.
                                       
            {self.__tip_section()}

            Make sure to do something else.
        """
            ),
        expected_output="Expected Output of this Agent(task_2_name) in simple words",
            agent=agent,
        )

I would like to create a PR on the following issue and Update few things like structure of passing Agent and Task using yaml file.