Sdílet prostřednictvím


Použití agenta jako funkčního nástroje

V tomto kurzu se dozvíte, jak používat agenta jako funkční nástroj, aby jeden agent mohl volat jiného agenta jako nástroj.

Požadavky

Pokyny k požadavkům a instalaci balíčků NuGet najdete v části Vytvoření a spuštění jednoduchého agenta v tomto kurzu.

Vytvoření a použití agenta jako nástroje funkcí

AIAgent můžete použít jako nástroj funkce tím, že zavoláte agenta .AsAIFunction() a poskytnete jej jako nástroj jinému agentovi. To vám umožní vytvářet agenty a vytvářet pokročilejší pracovní postupy.

Nejprve vytvořte nástroj funkce jako metodu jazyka C# a v případě potřeby ho vyzdobit popisy. Tento nástroj bude používat váš agent, který je vystavený jako funkce.

using System.ComponentModel;

[Description("Get the weather for a given location.")]
static string GetWeather([Description("The location to get the weather for.")] string location)
    => $"The weather in {location} is cloudy with a high of 15°C.";

Vytvořte nástroj AIAgent , který používá nástroj funkce.

using System;
using Azure.AI.OpenAI;
using Azure.Identity;
using Microsoft.Agents.AI;
using Microsoft.Extensions.AI;
using OpenAI;

AIAgent weatherAgent = new AzureOpenAIClient(
    new Uri("https://<myresource>.openai.azure.com"),
    new AzureCliCredential())
     .GetChatClient("gpt-4o-mini")
     .CreateAIAgent(
        instructions: "You answer questions about the weather.",
        name: "WeatherAgent",
        description: "An agent that answers questions about the weather.",
        tools: [AIFunctionFactory.Create(GetWeather)]);

Teď vytvořte hlavního agenta a nakonfigurujte weatherAgent jako nástroj funkce voláním .AsAIFunction(), abyste převedli weatherAgent na nástroj funkce.

AIAgent agent = new AzureOpenAIClient(
    new Uri("https://<myresource>.openai.azure.com"),
    new AzureCliCredential())
     .GetChatClient("gpt-4o-mini")
     .CreateAIAgent(instructions: "You are a helpful assistant who responds in French.", tools: [weatherAgent.AsAIFunction()]);

Vyvolejte hlavního agenta jako obvykle. Teď může volat agenta počasí jako nástroj a měl by reagovat ve francouzštině.

Console.WriteLine(await agent.RunAsync("What is the weather like in Amsterdam?"));

V tomto kurzu se dozvíte, jak používat agenta jako funkční nástroj, aby jeden agent mohl volat jiného agenta jako nástroj.

Požadavky

Informace o požadavcích a instalaci balíčků najdete v části Vytvoření a spuštění jednoduchého agenta v tomto kurzu.

Vytvoření a použití agenta jako nástroje funkcí

Jako nástroj funkce můžete použít ChatAgent tím, že vyvoláte .as_tool() u agenta a poskytnete ho jako nástroj jinému agentovi. To vám umožní vytvářet agenty a vytvářet pokročilejší pracovní postupy.

Nejprve vytvořte nástroj ve formě funkce, který bude používat váš agent a který je poskytován jako funkce.

from typing import Annotated
from pydantic import Field

def get_weather(
    location: Annotated[str, Field(description="The location to get the weather for.")],
) -> str:
    """Get the weather for a given location."""
    return f"The weather in {location} is cloudy with a high of 15°C."

Vytvořte ChatAgent, který používá nástroj funkce.

from agent_framework.azure import AzureOpenAIChatClient
from azure.identity import AzureCliCredential

weather_agent = AzureOpenAIChatClient(credential=AzureCliCredential()).create_agent(
    name="WeatherAgent",
    description="An agent that answers questions about the weather.",
    instructions="You answer questions about the weather.",
    tools=get_weather
)

Teď vytvořte hlavního agenta a nakonfigurujte weather_agent jako nástroj funkce voláním .as_tool(), abyste převedli weather_agent na nástroj funkce.

main_agent = AzureOpenAIChatClient(credential=AzureCliCredential()).create_agent(
    instructions="You are a helpful assistant who responds in French.",
    tools=weather_agent.as_tool()
)

Vyvolejte hlavního agenta jako obvykle. Teď může volat agenta počasí jako nástroj a měl by reagovat ve francouzštině.

result = await main_agent.run("What is the weather like in Amsterdam?")
print(result.text)

Při převodu agenta na nástroj můžete také přizpůsobit název nástroje, popis a název argumentu:

# Convert agent to tool with custom parameters
weather_tool = weather_agent.as_tool(
    name="WeatherLookup",
    description="Look up weather information for any location",
    arg_name="query",
    arg_description="The weather query or location"
)

main_agent = AzureOpenAIChatClient(credential=AzureCliCredential()).create_agent(
    instructions="You are a helpful assistant who responds in French.",
    tools=weather_tool
)

Další kroky