The MCP (Model Context Protocol), developed by Anthropic, is an open protocol that standardizes how applications provide context to LLMs.

​
MCP Tool

MCPTool allows you to consume external tools exposed via MCP protocol. See the MCP tool documentation for more information.

​
MCP Server

MCPServer allows you to expose your tools to external systems that support the Model Context Protocol (MCP) standard, enabling seamless integration with LLM tools ecosystems. Key benefits
  • Fast setup with minimal configuration
  • Support for multiple transport options
  • Register multiple tools on a single server
  • Custom server settings and instructions
from mcp.server.fastmcp.server import Settings

from beeai_framework.adapters.mcp.serve.server import MCPServer, MCPServerConfig
from beeai_framework.tools import tool
from beeai_framework.tools.types import StringToolOutput
from beeai_framework.tools.weather.openmeteo import OpenMeteoTool


@tool
def reverse_tool(word: str) -> StringToolOutput:
    """
    A tool that reverses a word
    """
    return StringToolOutput(result=word[::-1])


def main() -> None:
    # create a MCP server with custom config, register ReverseTool and OpenMeteoTool to the MCP server and run it
    MCPServer(config=MCPServerConfig(transport="sse", settings=Settings(port=8001))).register_many(
        [reverse_tool, OpenMeteoTool()]
    ).serve()


if __name__ == "__main__":
    main()
The MCP adapter uses the MCPServerConfig class to configure the MCP server: 
class MCPServerConfig(BaseModel):
"""Configuration for the MCPServer."""
transport: Literal["stdio", "sse"] = "stdio"  # Transport protocol (stdio or server-sent events)
name: str = "MCP Server"                     # Name of the MCP server
instructions: str | None = None              # Optional instructions for the server
settings: mcp_server.Settings[Any] = Field(default_factory=lambda: mcp_server.Settings())
Transport Options
  • stdio: Uses standard input/output for communication (default)
  • sse: Uses server-sent events over HTTP
Creating an MCP server is easy. You instantiate the MCPServer class with your configuration, register your tools, and then call serve() to start the server: 
from beeai_framework.adapters.mcp import MCPServer, MCPServerConfig
from beeai_framework.tools.weather import OpenMeteoTool

# Create an MCP server with default configuration
server = MCPServer()

# Register tools
server.register(OpenMeteoTool())

# Start serving
server.serve()
You can configure the server behavior by passing a custom configuration: 
from beeai_framework.adapters.mcp import MCPServer
from beeai_framework.tools.weather import OpenMeteoTool
from beeai_framework.tools.search.wikipedia import WikipediaTool
from beeai_framework.tools.search.duckduckgo import DuckDuckGoSearchTool

def main():
server = MCPServer()
server.register_many([
OpenMeteoTool(),
WikipediaTool(),
DuckDuckGoSearchTool()
])
server.serve()

if __name__ == "__main__":
main()
MCPTool lets you add MCP-compatible tools to any agent, see Tools documentation to learn more.