Use it from an agent framework

The address

https://plans.intoperson.com/mcp

LangChain

pip install "langchain[mcp]"

from langchain.mcp import MCPAdapter

async def main():
    async with MCPAdapter("https://plans.intoperson.com/mcp") as adapter:
        tools = await adapter.list_tools()

An http or https address connects over Streamable HTTP. The langchain.mcp namespace needs langchain[mcp] 1.4.0 or later and is in beta in that documentation.

Code shape as in https://docs.langchain.com/oss/python/langchain/mcp and https://docs.langchain.com/oss/python/migrate/langchain-mcp-adapters, seen 2026-10-20.

OpenAI Agents SDK

from agents import Agent, HostedMCPTool

agent = Agent(
    name="Plan checker",
    tools=[HostedMCPTool(tool_config={
        "type": "mcp",
        "server_label": "plan_finder",
        "server_url": "https://plans.intoperson.com/mcp",
        "require_approval": "never",
    })],
)

The hosted tool is called by OpenAI's side, so the address must be reachable from the internet.

Code shape as in https://openai.github.io/openai-agents-python/mcp/, seen 2026-10-20.

CrewAI

from crewai import Agent
from crewai.mcp import MCPServerHTTP

agent = Agent(
    role="Plan checker",
    goal="Check API plans against a workload",
    backstory="Reads plan prices and limits.",
    mcps=[
        MCPServerHTTP(
            url="https://plans.intoperson.com/mcp",
            streamable=True,
        ),
    ],
)

MCPServerHTTP with streamable=True is the remote-server form in the CrewAI structured configuration.

Code shape as in https://docs.crewai.com/en/mcp/overview, seen 2026-10-20.

Other clients

Chat clients and editors are in the guide; app builders in app builders.