Skip to main content
Your AgentOS will by default be exposed as an API. But you can also expose it as an MCP server. This is done by setting enable_mcp_server=True when creating your AgentOS instance: 
agent_os = AgentOS(enable_mcp_server=True)

​
Why use MCP?

The MCP protocol has become the industry standard to handle connecting AI applications with external tools and data sources. By exposing your AgentOS as an MCP server, external clients that can handle MCP-compatible applications will be able to connect to your AgentOS and interact with it.

​
Example

enable_mcp_example.py
from agno.agent import Agent
from agno.db.postgres import PostgresDb
from agno.models.anthropic import Claude
from agno.os import AgentOS
from agno.tools.duckduckgo import DuckDuckGoTools

# Setup the database
db = PostgresDb(db_url="postgresql+psycopg://ai:ai@localhost:5532/ai")

# Create your agents
web_research_agent = Agent(
    name="Web Research Agent",
    model=Claude(id="claude-sonnet-4-0"),
    db=db,
    tools=[DuckDuckGoTools()],
    markdown=True,
)

# Setup AgentOS with MCP enabled
agent_os = AgentOS(
    description="Example app with MCP enabled",
    agents=[web_research_agent],
    enable_mcp_server=True,  # This enables a LLM-friendly MCP server at /mcp
)

app = agent_os.get_app()

if __name__ == "__main__":
    # Your MCP server will be available at http://localhost:7777/mcp
    agent_os.serve(app="enable_mcp_example:app", reload=True)
Once enabled, your AgentOS will expose an MCP server via the /mcp endpoint. You can see a complete example here.

​
Available MCP Tools

When you expose your AgentOS as an MCP server, the following MCP tools will be available:

​
get_agentos_config

Get the configuration of the AgentOS

​
run_agent

Run the desired Agent
  • agent_id (str): The ID of the agent to run
  • message (str): The message to send to the agent

​
run_team

Run the desired Team
  • team_id (str): The ID of the team to run
  • message (str): The message to send to the team

​
run_workflow

Run the desired Workflow
  • workflow_id (str): The ID of the workflow to run
  • message (str): The message to send to the workflow

​
get_sessions_for_agent

Get the list of sessions for the desired Agent
  • agent_id (str): The ID of the agent to get the sessions for
  • db_id (str): The ID of the database to use
  • user_id (Optional[str]): The ID of the user to get the sessions for
  • sort_by (Optional[str]): The field to sort the sessions by. Defaults to created_at
  • sort_order (Optional[str]): The order to sort the sessions by. Defaults to desc

​
get_sessions_for_team

Get the list of sessions for the desired Team
  • team_id (str): The ID of the team to get the sessions for
  • db_id (str): The ID of the database to use
  • user_id (Optional[str]): The ID of the user to get the sessions for
  • sort_by (Optional[str]): The field to sort the sessions by. Defaults to created_at
  • sort_order (Optional[str]): The order to sort the sessions by. Defaults to desc

​
get_sessions_for_workflow

Get the list of sessions for the desired Workflow
  • workflow_id (str): The ID of the workflow to get the sessions for
  • db_id (str): The ID of the database to use
  • user_id (Optional[str]): The ID of the user to get the sessions for
  • sort_by (Optional[str]): The field to sort the sessions by. Defaults to created_at
  • sort_order (Optional[str]): The order to sort the sessions by. Defaults to desc

​
create_memory

Create a new user memory
  • db_id (str): The ID of the database to use
  • memory (str): The memory content to store
  • user_id (str): The user this memory is about
  • topics (Optional[list[str]]): The topics of the memory

​
get_memories_for_user

Get the list of memories for the given user
  • user_id (str): The ID of the user to get the memories for
  • db_id (Optional[str]): The ID of the database to use
  • sort_by (Optional[str]): The field to sort the memories by. Defaults to created_at
  • sort_order (Optional[str]): The order to sort the memories by. Defaults to desc

​
update_memory

Update the desired memory
  • db_id (str): The ID of the database to use
  • memory_id (str): The ID of the memory to update
  • memory (str): The memory content to store
  • user_id (str): The ID of the user to update the memory for

​
delete_memory

Delete the desired memory
  • db_id (str): The ID of the database to use
  • memory_id (str): The ID of the memory to delete
See a full example here.