[BUG]

[PremKumar135]

Description

Whenever i wanted to my agent to product a pydantic model output by giving a pydantic model to output_pydantic parameter, it is working somtimes or failing most of the time by giving this message
Failed to add to long term memory: Failed to convert text into a Pydantic model due to error: No module named 'google.genai' irrespective of success or failure. Because of this i cannot use that task in the next flow.

Steps to Reproduce

Create a simple agent and pydantic model and in agent , give that pydantic model as a output_pydantic, we can see the error

Expected behavior

It should work without any error message and the next agent should understand that it is a pydantic model and the second agent should be able convert that into a string automatically and and work from there. Or we should have the option of what key value pair from the pydantic model of the first agent can be given as a input to second agent.

Screenshots/Code snippets

class ResponseFormat(BaseModel):
string : str = Field(description = 'string needs to manitained')

class FirstFlowCrew:

def first_crew(self, param):
    first_agent = Agent(role = "role mentioned",
                            goal = dedent( f"""Goals mentioned"""),
                            backstory = dedent("""Back story mentioned"""),
                            tools=[tool mentioned],
                            verbose=True,
                            allow_delegation=True,
                            memory = True
                        )

   task_1 = Task(description = "Task description mentioned",
                               expected_output = "expected output mentioned.",
                               agent = first_agent
                            )
    

    second_agent = Agent(role='role mentioned',
                               goal = dedent("""goal mentioned"""),
                                backstory= "back story mentioned",
                                
                            )
    
    task_2 = Task(description = "task mentioned",
                             expected_output = "expected output metioned",
                             agent = second_agent,
                            output_pydantic=ResponseFormat
                            )

    first_crew = Crew(agents = [first_agent, second_agent], 
                tasks = [task_1, task_2], 
                process = 'sequential',
                verbose = True,
                memory=True
                )
    
    result = first_crew.kickoff()
    return result

Operating System

Other (specify in additional context)

Python Version

3.12

crewAI Version

0.108.0

crewAI Tools Version

0.38.1

Virtual Environment

Venv

Evidence

Possible Solution

None

Additional context

None

[Vidit-Ostwal]

Can you try pip install google-generativeai