Anyone managed to use mcp server with ollama model provider?

smrati_katiyar · June 30, 2025, 6:20pm

Hey everyone ,
I am trying to use a local llm while using mcp servers in crew ai
i can see in verbose log that llm is able to identify correct tool and able to run it and receive output from tool but after that step it throws an APIConnectionError

Here is my code snippet

from crewai import Agent, Task, Crew, LLM, Process
from crewai_tools import MCPServerAdapter
from mcp import StdioServerParameters
import os

from crewai import LLM

server_params=StdioServerParameters(
    command="uv", # Or your python3 executable i.e. "python3"
    args=["run", "/home/load_testing_stdio/server.py"],
)

with MCPServerAdapter(server_params) as tools:
    print(f"Available tools from Stdio MCP server: {[tool.name for tool in tools]}")

    my_llm = LLM(
        model="ollama/llama3.2",
        base_url="http://localhost:11434",
        streaming=True
    )


    # Example: Using the tools from the Stdio MCP server in a CrewAI Agent
    agent = Agent(
        role="Hash Calculator",
        goal="compute hash of a given string",
        backstory="You are a hash calculator, you compute hash of a given string",
        tools=tools,
        verbose=True,
        allow_delegation=False,
        llm=my_llm,
    )
    task = Task(
        description="Compute hash for the string 'hello world'",
        expected_output="return hash of specified string.",
        agent=agent,
    )
    crew = Crew(
        agents=[agent],
        tasks=[task],
        verbose=True,
        process=Process.sequential,
    )
    result = crew.kickoff()
    # print(result)

Here is the console output

/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/pydantic/fields.py:1093: PydanticDeprecatedSince20: Using extra keyword arguments on `Field` is deprecated and will be removed. Use `json_schema_extra` instead. (Extra keys: 'required'). Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.11/migration/
  warn(
[06/30/25 11:13:16] INFO     Processing request of type ListToolsRequest                                                                            server.py:619
                    INFO     Processing request of type ListToolsRequest                                                                            server.py:619
/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/pydantic/fields.py:1093: PydanticDeprecatedSince20: Using extra keyword arguments on `Field` is deprecated and will be removed. Use `json_schema_extra` instead. (Extra keys: 'items', 'anyOf', 'enum', 'properties'). Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.11/migration/
  warn(
Available tools from Stdio MCP server: ['generate_md5_hash', 'count_characters', 'get_first_half']
╭─────────────────────────────────────────────────────────────────── Crew Execution Started ────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  Crew Execution Started                                                                                                                                       │
│  Name: crew                                                                                                                                                   │
│  ID: c6e83991-152f-43ff-a696-8c02b5ccbe8e                                                                                                                     │
│  Tool Args:                                                                                                                                                   │
│                                                                                                                                                               │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯

🚀 Crew: crew
└── 📋 Task: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d
    Status: Executing Task...
╭────────────────────────────────────────────────────────────────────── 🤖 Agent Started ───────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  Agent: Hash Calculator                                                                                                                                       │
│                                                                                                                                                               │
│  Task: Compute hash for the string 'hello world'                                                                                                              │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯

🚀 Crew: crew
└── 📋 Task: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d
    Status: Executing Task...
🚀 Crew: crew
└── 📋 Task: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d
    Status: Executing Task...
    └── 🔧 Used generate_md5_hash (1)
╭─────────────────────────────────────────────────────────────────── 🔧 Agent Tool Execution ───────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  Agent: Hash Calculator                                                                                                                                       │
│                                                                                                                                                               │
│  Thought: Thought: I need to compute the hash of the input string 'hello world'                                                                               │
│                                                                                                                                                               │
│  Using Tool: generate_md5_hash                                                                                                                                │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭───────────────────────────────────────────────────────────────────────── Tool Input ──────────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  "{\"input_str\": \"hello world\"}"                                                                                                                           │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭───────────────────────────────────────────────────────────────────────── Tool Output ─────────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  5eb63bbbe01eeed093cb22bb8f5acdc3                                                                                                                             │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯





🚀 Crew: crew
└── 📋 Task: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d
    Status: Executing Task...
    ├── 🔧 Used generate_md5_hash (1)
    └── ❌ LLM Failed


 An unknown error occurred. Please check the details below.

🚀 Crew: crew
└── 📋 Task: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d
    Assigned to: Hash Calculator
    Status: ❌ Failed
    ├── 🔧 Used generate_md5_hash (1)
    └── ❌ LLM Failed
╭──────────────────────────────────────────────────────────────────────── Task Failure ─────────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  Task Failed                                                                                                                                                  │
│  Name: 722ec58f-28ff-4a12-bcaa-c6055bfaf11d                                                                                                                   │
│  Agent: Hash Calculator                                                                                                                                       │
│  Tool Args:                                                                                                                                                   │
│                                                                                                                                                               │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯

╭──────────────────────────────────────────────────────────────────────── Crew Failure ─────────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                               │
│  Crew Execution Failed                                                                                                                                        │
│  Name: crew                                                                                                                                                   │
│  ID: c6e83991-152f-43ff-a696-8c02b5ccbe8e                                                                                                                     │
│  Tool Args:                                                                                                                                                   │
│  Final Output:                                                                                                                                                │
│                                                                                                                                                               │
│                                                                                                                                                               │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯

Traceback (most recent call last):
    response = base_llm_http_handler.completion(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    data = provider_config.transform_request(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    modified_prompt = ollama_pt(model=model, messages=messages)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    tool_calls = messages[msg_i].get("tool_calls")
                 ~~~~~~~~^^^^^^^
IndexError: list index out of range

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/smrati/smrati_linux_subsystem/learning/weather/crewai_mcp_server_integration/stdio_integration.py", line 44, in <module>
    result = crew.kickoff()
             ^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/crew.py", line 659, in kickoff
    result = self._run_sequential_process()
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/crew.py", line 768, in _run_sequential_process
    return self._execute_tasks(self.tasks)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/crew.py", line 871, in _execute_tasks
    task_output = task.execute_sync(
                  ^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/task.py", line 354, in execute_sync
    return self._execute_core(agent, context, tools)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/task.py", line 502, in _execute_core
    raise e  # Re-raise the exception after emitting the event
    ^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/task.py", line 418, in _execute_core
    result = agent.execute_task(
             ^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agent.py", line 435, in execute_task
    raise e
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agent.py", line 411, in execute_task
    result = self._execute_without_timeout(task_prompt, task)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agent.py", line 507, in _execute_without_timeout
    return self.agent_executor.invoke(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agents/crew_agent_executor.py", line 125, in invoke
    raise e
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agents/crew_agent_executor.py", line 114, in invoke
    formatted_answer = self._invoke_loop()
                       ^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agents/crew_agent_executor.py", line 210, in _invoke_loop
    raise e
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/agents/crew_agent_executor.py", line 157, in _invoke_loop
    answer = get_llm_response(
             ^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/utilities/agent_utils.py", line 160, in get_llm_response
    raise e
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/utilities/agent_utils.py", line 151, in get_llm_response
    answer = llm.call(
             ^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/llm.py", line 956, in call
    return self._handle_non_streaming_response(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/smrati/smrati_linux_subsystem/learning/weather/.venv/lib/python3.11/site-packages/crewai/llm.py", line 768, in _handle_non_streaming_response
    
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    raise e
    result = original_function(*args, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    raise exception_type(
          ^^^^^^^^^^^^^^^
    raise e
    raise APIConnectionError(
.APIConnectionError:

as you can see llm is able to fetch tools from mcp server, capture output from tool but after that crewai throws an error

smrati_katiyar · June 30, 2025, 6:27pm

i am using crewai version
crewai 0.134.0
crewai-tools 0.48.0

I have tried it with following LLMs
qwen3:8b
deepseek-r1:8b
llama3.2:3b

smrati_katiyar · July 25, 2025, 10:36pm

Hey i have managed to get it working.
I have found patchy solution found in this thread.

github.com/crewAIInc/crewAI

[BUG] IndexError: list index out of range

opened 08:06AM - 21 May 25 UTC

closed 06:48AM - 22 May 25 UTC

BurningSpy

bug

### Description Hi, this might be the wrong place to ask but the deeplearning….ai page doesnt have a Q&A or anywhere to ask. I am currently doing the second tutorial for "Practical Multi AI Agents and Advanced Use Cases" with crewai. In the tutorial we are using Trello, but since I don't use Trello I wanted to try out making a custom Tool for Asana instead. The tool itself seems to run fine. I give my workspace and project as inputs and the llm correctly calls the tool with the inputs and gets a response. However, the llm seems to not be able to handle the response. I let the Tool Format the list of tasks that i am getting from the project into a string to make it easier to understand for the model, but I am getting an IndexError as follows: # Agent: Data Collection Specialist ## Task: Create an initial understanding of the project, its main tasks and the team working on it. Use the Asana Data Fetcher tool to gather data from the Asana project. 🤖 Agent: Data Collection Specialist Status: In Progress └── 🧠 Thinking... 🤖 Agent: Data Collection Specialist Status: In Progress 🤖 Agent: Data Collection Specialist Status: In Progress DEBUG: BoardDataFetcherTool _run kwargs: {'workspace_name': 'myworkspace', 'project_name': 'myproject'} # Agent: Data Collection Specialist ## Thought: Action: Asana Board Data Fetcher ## Using tool: Asana Board Data Fetcher ## Tool Input: "{\"workspace_name\": \"myworkspace\", \"project_name\": \"myproject\"}" ## Tool Output: ### Asana Tasks: - **Check out freqtrade options** (GID: 1207913426279748) - **Check if mongodb might be an option for storing trades** (GID: 1207913426279737) - **Reset alarms ** (GID: 1207913426279733) - **Make helper ** (GID: 1207913426279729) - **Remove DBHelper from usage, but keep for later** (GID: 1208664419290176) - **refactor main ** (GID: 1208664419290180) - **Improve calculation** (GID: 1208664419290184) - **Improve saving of inf** (GID: 1208664419290186) - **Setup proper retrieval and analysis with db** (GID: 1207913426279746) 🤖 Agent: Data Collection Specialist Status: In Progress └── 🧠 Thinking... Provider List: https://docs.litellm.ai/docs/providers 🚀 Crew: crew └── 📋 Task: f648f16c-cd0a-4e90-b64a-59acbf76cac4 Status: Executing Task... └── 🤖 Agent: Data Collection Specialist Status: In Progress └── ❌ LLM Failed ╭────────────────────────────────────────────────────────────────────────────────── LLM Error ──────────────────────────────────────────────────────────────────────────────────╮│ ││ ❌ LLM Call Failed │ │ Error: litellm.APIConnectionError: list index out of range ││ Traceback (most recent call last): ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\main.py", line 2904, in completion ││ response = base_llm_http_handler.completion( ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\llms\custom_httpx\llm_http_handler.py", line 270, in completion ││ data = provider_config.transform_request( ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\llms\ollama\completion\transformation.py", line 322, in transform_request ││ modified_prompt = ollama_pt(model=model, messages=messages) ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\litellm_core_utils\prompt_templates\factory.py", line 229, in ollama_pt ││ tool_calls = messages[msg_i].get("tool_calls") ││ ~~~~~~~~^^^^^^^ ││ IndexError: list index out of range ││ ││ │╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ I am using Ollama for the agent and "llama3.1:8b" as model (I've used it in previous tutorials where it worked fine and was able to use tools as well) ### Steps to Reproduce This the custom Tool that I setup class AsanaInput(BaseModel): """Input for fetching data from an Asana board.""" workspace_name: str = Field(..., description="Name of the Asana workspace") project_name: str = Field(..., description="Name of the project within the workspace") class BoardDataFetcherTool(BaseTool): name: str = "Asana Board Data Fetcher" description: str = "Fetches tasks from an Asana project using workspace name and project name" args_schema: Type[BaseModel] = AsanaInput # Add this line access_token: str = os.environ['ASANA_ACCESS_TOKEN'] def _get_workspace_gid(self, workspace_name: str) -> str: """Get workspace GID from workspace name""" url = "https://app.asana.com/api/1.0/workspaces" headers = { 'Authorization': f'Bearer {self.access_token}', } response = requests.get(url, headers=headers) if response.status_code == 200: workspaces = response.json().get('data', []) for workspace in workspaces: if workspace['name'].lower() == workspace_name.lower(): return workspace['gid'] return None def _get_project_gid(self, workspace_gid: str, project_name: str) -> str: """Get project GID from project name within a workspace""" url = f"https://app.asana.com/api/1.0/projects" headers = { 'Authorization': f'Bearer {self.access_token}', } params = { 'workspace': workspace_gid } response = requests.get(url, headers=headers, params=params) if response.status_code == 200: projects = response.json().get('data', []) for project in projects: if project['name'].lower() == project_name.lower(): return project['gid'] return None def _run(self, **kwargs) -> dict: print("DEBUG: BoardDataFetcherTool _run kwargs:", kwargs) # Add this line """ Fetch all tasks from the specified workspace and project. Args: kwargs: Dictionary containing workspace_name and project_name Returns: dict: JSON response containing task data """ workspace_name = kwargs.get('workspace_name') project_name = kwargs.get('project_name') if not workspace_name or not project_name: return {"error": "Missing workspace_name or project_name"} try: # Get workspace GID workspace_gid = self._get_workspace_gid(workspace_name) if not workspace_gid: return {"error": f"Workspace '{workspace_name}' not found"} # Get project GID project_gid = self._get_project_gid(workspace_gid, project_name) if not project_gid: return {"error": f"Project '{project_name}' not found in workspace '{workspace_name}'"} # Get tasks url = f"https://app.asana.com/api/1.0/projects/{project_gid}/tasks" headers = { 'Authorization': f'Bearer {self.access_token}', } response = requests.get(url, headers=headers) if response.status_code == 200: tasks = response.json().get("data", []) if not tasks: return "No tasks found in this project." output = "### Asana Tasks:\n" for t in tasks: output += f"- **{t['name']}** (GID: {t['gid']})\n" return output else: return f"Failed to fetch tasks. Status code: {response.status_code}" except Exception as e: return {"error": f"An error occurred: {str(e)}"} ### Expected behavior I'd expect the llm to accept the returned string and keep working with the new knowledge ### Screenshots/Code snippets # Creating Crew crew = Crew( agents=[ data_collection_agent, # analysis_agent ], tasks=[ data_collection, # data_analysis, # report_generation ], verbose=True, embedder={ "provider": "ollama", "config": { "model": "nomic-embed-text:latest" } } ) # The given Python dictionary inputs = { 'workspace_name': workspace_name, 'project_name': project_name } # Run the crew result = crew.kickoff( inputs=inputs ) ### Operating System Windows 10 ### Python Version 3.11 ### crewAI Version 0.119.0 ### crewAI Tools Version 0.45.0 ### Virtual Environment Venv ### Evidence ╭────────────────────────────────────────────────────────────────────────────────── LLM Error ──────────────────────────────────────────────────────────────────────────────────╮│ ││ ❌ LLM Call Failed │ │ Error: litellm.APIConnectionError: list index out of range ││ Traceback (most recent call last): ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\main.py", line 2904, in completion ││ response = base_llm_http_handler.completion( ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\llms\custom_httpx\llm_http_handler.py", line 270, in completion ││ data = provider_config.transform_request( ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\llms\ollama\completion\transformation.py", line 322, in transform_request ││ modified_prompt = ollama_pt(model=model, messages=messages) ││ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ││ File "C:\Users\username\AppData\Roaming\Python\Python311\site-packages\litellm\litellm_core_utils\prompt_templates\factory.py", line 229, in ollama_pt ││ tool_calls = messages[msg_i].get("tool_calls") ││ ~~~~~~~~^^^^^^^ ││ IndexError: list index out of range ││ ││ │╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ ### Possible Solution I'm assuming that there is something wrong with my setup, but looking at the documentation it seems to be correct. I'm unsure what to do. I've tried debugging and hit a wall. ### Additional context I use Ollama with llama3.1:8b the agent yaml: data_collection_agent: role: > Data Collection Specialist goal: > Gather all relevant data from the asana project in the given workspace. workspace name is {workspace_name} project name is {project_name} backstory: > You are responsible for ensuring that all project data is collected accurately. allow_delegation: false verbose: true task yaml: data_collection: description: > Create an initial understanding of the project, its main tasks and the team working on it. Use the Asana Data Fetcher tool to gather data from the Asana project. expected_output: > A full blown report on the project, including its main features, the team working on it, and any other relevant information from the Asana board.

So basically it is not a crewai bug but a litellm bug
To fix this bug you need to go to this path in litellm package
.venv/lib/python3.11/site-packages/litellm/litellm_core_utils/prompt_templates/factory.py

Search for this comment

## MERGE CONSECUTIVE ASSISTANT CONTENT ##

And do these two code changes

After this you will be able to run your crews succesfully using MCP as tools with Ollama as LLM provider

I have tested it on version

crewai                                   0.134.0
crewai-tools                             0.48.0

Here is my sample code snippet

from crewai import Agent, Task, Crew, LLM, Process
from crewai_tools import MCPServerAdapter
from mcp import StdioServerParameters
import os
# from langchain.llms import Ollama
from langchain_openai import ChatOpenAI

from crewai import LLM

server_params=StdioServerParameters(
    command="uv", # Or your python3 executable i.e. "python3"
    args=["run", "/home/jon/doe/load_testing_stdio/server.py"],
)

with MCPServerAdapter(server_params) as tools:
    print(f"Available tools from Stdio MCP server: {[tool.name for tool in tools]}")

    # my_llm = LLM(
    #     model="ollama/llama3.2",
    #     base_url="http://localhost:11434",
    #     streaming=True
    # )

    # my_llm = Ollama(model="ollama/llama3.2")

    my_llm = ChatOpenAI(
        model="ollama/llama3.2",
        base_url="http://localhost:11434",
        api_key="sk-ollama",
        stream=True
    )


    # Example: Using the tools from the Stdio MCP server in a CrewAI Agent
    agent = Agent(
        role="Hash Calculator",
        goal="compute hash of a given string",
        backstory="You are a hash calculator, you compute hash of a given string",
        tools=tools,
        verbose=True,
        allow_delegation=False,
        llm=my_llm,
    )
    task = Task(
        description="Compute hash for the string 'hello world'",
        expected_output="return hash of specified string.",
        agent=agent,
        verbose=True
    )
    crew = Crew(
        agents=[agent],
        tasks=[task],
        verbose=True,
        process=Process.sequential,
    )
    result = crew.kickoff()
    print(result)

Here is screenshot of successful completion of my crew

I case you want to quickly test it , here is my MCP server

from typing import Any
import hashlib
import time
from loguru import logger
from mcp.server.fastmcp import FastMCP

# Initialize FastMCP server
mcp = FastMCP("mcp-1")

@mcp.tool()
def generate_md5_hash(input_str: str) -> str:
    # Create an md5 hash object
    logger.info(f"Generating MD5 hash for: {input_str}")
    md5_hash = hashlib.md5()

    # add forced delay
    time.sleep(20)
    
    # Update the hash object with the bytes of the input string
    md5_hash.update(input_str.encode('utf-8'))
    
    # Return the hexadecimal representation of the digest
    return md5_hash.hexdigest()

@mcp.tool()
def count_characters(input_str: str) -> int:
    # Count number of characters in the input string
    logger.info(f"Counting characters in: {input_str}")
    return len(input_str)


@mcp.tool()
def get_first_half(input_str: str) -> str:
    # Calculate the midpoint of the string
    logger.info(f"Getting first half of: {input_str}")
    midpoint = len(input_str) // 2
    
    # Return the first half of the string
    return input_str[:midpoint]


if __name__ == "__main__":
    # Initialize and run the server
    mcp.run(transport='stdio')

system · July 26, 2025, 10:37pm

This topic was automatically closed 24 hours after the last reply. New replies are no longer allowed.

Topic		Replies	Views
LLM failed after calling a tool from a user-defined MCP server CrewAI Community Support tools_issues	7	136	July 18, 2025
Agents can discover & use tools hosted on Model Context Protocol (MCP) Server General tools_issues	7	2556	April 30, 2025
Mcp server tool - config, adapter example - in CrewAI cli setup CrewAI Community Support	6	577	May 25, 2025
CrewAI MCP Integration live in Docs! 🚀 Announcements community , release , news	0	356	May 21, 2025
Connecting Ollama with crewai Crews crewai	14	9451	July 19, 2025

Anyone managed to use mcp server with ollama model provider?

Related topics