在 MCP 伺服器中使用 LLM 工具
你已經建立了一個 MCP 伺服器,裡面包含一個能以最新匯率進行幣別換算的工具。把它與 LLM 整合後,模型就能精準回答關於各種幣別與匯率的問題——這是它在預設狀態下做不到的。
這裡已為你提供大部分程式碼,因為重點是理解工作流程而非語法細節。
本練習屬於課程
Model Context Protocol(MCP)入門
練習說明
- 將使用者查詢(
user_query)與格式化的工具清單(anthropic_tools)送給 Claude。 - 使用從工具使用區塊擷取的名稱與引數,呼叫 LLM 所選擇的 MCP 工具。
- 將結果(
result)回傳給 Claude,取得最終回應。
動手互動練習
試著完成這個範例程式碼,體驗一下這個練習。
from mcp import ClientSession, StdioServerParameters
from mcp.client.stdio import stdio_client
async def call_anthropic_llm(user_query: str):
"""Call Claude with MCP tools."""
print(f"\nUser: {user_query}\n")
mcp_tools = await get_tools_from_mcp()
anthropic_tools = []
for tool in mcp_tools:
anthropic_tool = {
"name": tool.name,
"description": tool.description or "",
"input_schema": tool.inputSchema,
}
anthropic_tools.append(anthropic_tool)
# Send the user query and formatted tools to the LLM
client = AsyncAnthropic(api_key="")
response = await client.messages.create(
model="claude-sonnet-4-6",
max_tokens=1024,
messages=____,
tools=____,
)
if response.stop_reason == "tool_use":
tool_use = next(block for block in response.content if block.type == "tool_use")
name = tool_use.name
args = tool_use.input
print(f"Model decided to call: {name}")
print(f"Arguments: {args}\n")
# Call the MCP tool
result = await call_mcp_tool(____, ____)
# Send the result back to Claude for the final response
followup = await client.messages.create(
model="claude-sonnet-4-6",
max_tokens=1024,
messages=[
{"role": "user", "content": user_query},
{"role": "assistant", "content": response.content},
{
"role": "user",
"content": [
{
"type": "tool_result",
"tool_use_id": tool_use.id,
"content": ____,
}
],
},
],
tools=anthropic_tools,
)
final_text = next((block.text for block in followup.content if block.type == "text"), None)
if final_text:
print(f"\nAssistant: {final_text}")
return str(final_text)
else:
print("No follow-up message from model.")
else:
text = next((block.text for block in response.content if block.type == "text"), "")
print(f"\nAssistant: {text}")
return str(text)
if __name__ == "__main__":
asyncio.run(call_anthropic_llm("How much is 250 US dollars in euros?"))