将 LlamaIndex 智能体工具作为 MCP 工具使用¶

我们在 LlamaHub 中拥有数十种智能体工具，它们都可以立即作为 MCP 工具使用！本笔记本将以 Notion 工具为例，具体展示如何实现这一功能。

首先我们安装工具和 MCP 服务器：

In [ ]:

!pip install llama-index-tools-notion mcp fastmcp

!pip install llama-index-tools-notion mcp fastmcp

引入依赖项：

In [ ]:

# Import dependencies for Model Context Protocol (MCP) fastMCP server
from typing import Any, Dict, List, Optional
from fastmcp import FastMCP

print("MCP fastMCP server dependencies imported successfully!")

# Import dependencies for Model Context Protocol (MCP) fastMCP server from typing import Any, Dict, List, Optional from fastmcp import FastMCP print("MCP fastMCP server dependencies imported successfully!")

MCP fastMCP server dependencies imported successfully!

使用 API 密钥实例化我们的工具：

In [ ]:

# Import and configure LlamaIndex Notion Tool Spec
from llama_index.tools.notion import NotionToolSpec

notion_token = "xxxx"
tool_spec = NotionToolSpec(integration_token=notion_token)

# Import and configure LlamaIndex Notion Tool Spec from llama_index.tools.notion import NotionToolSpec notion_token = "xxxx" tool_spec = NotionToolSpec(integration_token=notion_token)

让我们看看有哪些可用工具：

In [ ]:

tools = tool_spec.to_tool_list()

for i, tool in enumerate(tools):
    print(f"Tool {i+1}: {tool.metadata.name}")

tools = tool_spec.to_tool_list() for i, tool in enumerate(tools): print(f"Tool {i+1}: {tool.metadata.name}")

Tool 1: load_data
Tool 2: search_data

现在我们创建并配置 fastMCP 服务器，并注册每个工具：

In [ ]:

mcp_server = FastMCP("MCP Agent Tools Server")

# Register the tools from the Notion ToolSpec
for tool in tools:
    mcp_server.tool(
        name=tool.metadata.name, description=tool.metadata.description
    )(tool.real_fn)

mcp_server = FastMCP("MCP Agent Tools Server") # Register the tools from the Notion ToolSpec for tool in tools: mcp_server.tool( name=tool.metadata.name, description=tool.metadata.description )(tool.real_fn)

MCP Server configured with tools

现在我们可以运行完整的 MCP 服务器及其工具了！

In [ ]:

await mcp_server.run_async(transport="streamable-http")

await mcp_server.run_async(transport="streamable-http")

[06/27/25 16:27:47] INFO     Starting MCP server 'MCP Agent Tools Server' with transport             server.py:1358
                             'streamable-http' on http://127.0.0.1:8000/mcp/                                       

INFO:     Started server process [24668]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)

INFO:     127.0.0.1:54201 - "POST /mcp/ HTTP/1.1" 200 OK
INFO:     127.0.0.1:54201 - "POST /mcp/ HTTP/1.1" 202 Accepted
INFO:     127.0.0.1:54203 - "GET /mcp/ HTTP/1.1" 200 OK
INFO:     127.0.0.1:54209 - "POST /mcp/ HTTP/1.1" 200 OK

INFO:     Shutting down
ERROR:    ASGI callable returned without completing response.
ERROR:    Cancel 0 running task(s), timeout graceful shutdown exceeded
INFO:     Waiting for application shutdown.
INFO:     Application shutdown complete.
INFO:     Finished server process [24668]