[BUG] Training fails due to 'Manager agent should not have tools'

[manni07]

Description

The crew is running, doing its job and almost finished, final stage with last agent I get the following error message:

'
[2025-02-14 20:49:46][WARNING]: Manager agent should not have tools

[2025-02-14 20:49:46][ERROR]: Training failed: Manager agent should not have tools
Traceback (most recent call last):

...
'

Steps to Reproduce

1. Start crew
2. Wait up to the end
3. Error message appears

Expected behavior

Finalizing task

Screenshots/Code snippets

All agents are defined like this:

@agent def jury(self) -> Agent: return Agent( config=self.agents_config['jury'], verbose=True, allow_delegation=True, temperature=0.7, memory=True, respect_context_window=True, )

The crew is

` @crew
def crew(self) -> Crew:
"""Creates the Mycrew crew"""

	return Crew(
		agents=self.agents,
		tasks=self.tasks, 
		verbose=True,
		manager_llm="gpt-4o",
		memory=True,
		process=Process.hierarchical
	)

`

Operating System

Ubuntu 24.04

Python Version

3.12

crewAI Version

0.102.0

crewAI Tools Version

latest

Virtual Environment

Venv

Evidence

`[2025-02-14 20:49:46][WARNING]: Manager agent should not have tools

[2025-02-14 20:49:46][ERROR]: Training failed: Manager agent should not have tools
Traceback (most recent call last):`

Possible Solution

None

Additional context

No available documentation

[Vidit-Ostwal]

Can you share the entire code once.

[manni07]

Here the entire crew.py:

`
from crewai import Agent, Crew, Process, Task
from crewai.project import CrewBase, agent, crew, task
import logging

logger = logging.getLogger(name)

@crewbase
class Mycrew():
"""Mycrew crew"""

def __init__(self):
	logger.debug("Initializing Mycrew")
	super().__init__()

agents_config = 'config/agents.yaml'
tasks_config = 'config/tasks.yaml'

@agent
def researcher(self) -> Agent:
	return Agent(
		config=self.agents_config['researcher'],
		verbose=True,
		allow_delegation=False
	)

@agent
def assistant(self) -> Agent:
	return Agent(
        config=self.agents_config['assistant'],
        verbose=True,
        allow_delegation=False,
        temperature=0.7,
        memory=False
	)

@agent
def publisher(self) -> Agent:
	return Agent(
		config=self.agents_config['publisher'],
		verbose=True,
		allow_delegation=False,
		temperature=0.7,
		memory=False
	)

@task
def research_task(self) -> Task:
	return Task(
		config=self.tasks_config['research_task']
	)

@task
def assistant_task(self) -> Task:
	return Task(
		config=self.tasks_config['assistant_task']
	)

@task
def publisher_task(self) -> Task:
	return Task(
		config=self.tasks_config['publisher_task'],
		agent=self.publisher(),
		memory=True,
		cache=True,
		output_file='report.md'
	)

@crew
def crew(self) -> Crew:
	"""Creates the Mycrew crew"""

	crew_instance = Crew(
		agents=self.agents, # Automatically created by the @agent decorator
		tasks=self.tasks, # Automatically created by the @task decorator
		verbose=True,
		manager_llm="gpt-4o",
		memory=True,
		process=Process.hierarchical
	)
	logger.info("Crew created successfully")
	return crew_instance

`

[manni07]

Maybe of interest, it is repeatedly asking for a tool even though I did not specify:

`Processing training feedback.

Error parsing LLM output, agent will retry: I did it wrong. Invalid Format: I missed the 'Action:' after 'Thought:'. I will do right next, and don't use a tool I have already used.

If you don't need to use any more tools, you must give your best complete final answer, make sure it satisfies the expected criteria, use the EXACT format below:

Thought: I now can give a great answer
Final Answer: my best complete final answer to the task.`