Azure Agentic AI自定义函数调用失败：找不到fetch_weather函数

问题根源 #

报错Function 'fetch_weather' not found是因为函数注册到工具集的方式不符合Azure AI Agents的API要求，你当前用Set包裹函数并传入FunctionTool的写法有误，导致工具无法正确关联到目标函数。

具体修复步骤 #

1. 修正函数注册逻辑 #

不要用集合存储函数，改用以下两种正确方式之一：

方式一：直接包装单个函数（简洁版）

from azure.ai.agents import FunctionTool, ToolSet
import json

def fetch_weather(location: str) -> str:
    """
    Fetches the weather information for the specified location.

    :param location: The location to fetch weather for.
    :return: Weather information as a JSON string.
    """
    mock_weather_data = {"New York": "Sunny, 25°C", "London": "Cloudy, 18°C", "Tokyo": "Rainy, 22°C"}
    weather = mock_weather_data.get(location, "Weather data not available for this location.")
    return json.dumps({"weather": weather})

# 直接用FunctionTool包装目标函数
weather_tool = FunctionTool(fetch_weather)
toolset = ToolSet()
toolset.add(weather_tool)

方式二：显式定义函数元数据（适合复杂场景）

如果需要更精准控制函数的描述和参数规则，用FunctionDefinition显式声明：

from azure.ai.agents import FunctionTool, ToolSet
from azure.ai.agents.models import FunctionDefinition
import json

def fetch_weather(location: str) -> str:
    mock_weather_data = {"New York": "Sunny, 25°C", "London": "Cloudy, 18°C", "Tokyo": "Rainy, 22°C"}
    weather = mock_weather_data.get(location, "Weather data not available for this location.")
    return json.dumps({"weather": weather})

# 定义函数元数据
weather_function_def = FunctionDefinition(
    name="fetch_weather",
    description="Fetches the weather information for the specified location.",
    parameters={
        "type": "object",
        "properties": {
            "location": {
                "type": "string",
                "description": "City name to get weather for, e.g., 'New York'"
            }
        },
        "required": ["location"]
    }
)

# 绑定元数据与实现函数
weather_tool = FunctionTool.from_definition(weather_function_def, fetch_weather)
toolset = ToolSet()
toolset.add(weather_tool)

2. 修复Docstring格式 #

原函数docstring中param的写法错误，去掉参数类型的括号包裹，改为:param location: 描述，否则会干扰工具对参数的解析逻辑。

3. 升级依赖包 #

确保使用最新稳定版的Azure AI Agents SDK，避免版本兼容问题：

pip install --upgrade azure-ai-agents azure-identity

验证注意事项 #

• 代理初始化时必须正确传入toolset参数
• 测试时使用明确的输入（如“查询纽约的天气”），避免模糊表述导致参数解析失败

内容的提问来源于stack exchange，提问作者Tech_fusion