Compartilhar via


Explorando o kernel semântico AzureAIAgent

Importante

Esse recurso está em estágio experimental. As funcionalidades nesta fase ainda estão em desenvolvimento e estão sujeitas a alterações antes de avançarem para a fase de pré-visualização ou candidato a lançamento.

A documentação detalhada da API relacionada a esta discussão está disponível em:

A documentação atualizada da API do Semantic Kernel em Python estará disponível em breve.

No momento, os agentes não estão disponíveis em Java.

O que é um AzureAIAgent?

Um AzureAIAgent é um agente especializado dentro da estrutura do Semantic Kernel, projetado para fornecer capacidades conversacionais avançadas com integração perfeita de ferramentas. Ele automatiza a chamada de ferramenta, eliminando a necessidade de análise manual e invocação. O agente também gerencia com segurança o histórico de conversas usando threads, reduzindo a sobrecarga de manutenção do estado. Além disso, o AzureAIAgent dá suporte a uma variedade de ferramentas internas, incluindo recuperação de arquivo, execução de código e interação de dados por meio do Bing, Azure AI Search, Azure Functions e OpenAPI.

Para utilizar um AzureAIAgent, é necessário usar um Projeto do Azure AI Foundry. Os artigos a seguir fornecem uma visão geral do Azure AI Foundry, como criar e configurar um projeto e o serviço do agente:

Preparando seu ambiente de desenvolvimento

Para continuar desenvolvendo um AzureAIAgent, configure seu ambiente de desenvolvimento com os pacotes apropriados.

Adicione o pacote de Microsoft.SemanticKernel.Agents.AzureAI ao seu projeto:

dotnet add package Microsoft.SemanticKernel.Agents.AzureAI --prerelease

Talvez você também queira incluir o pacote Azure.Identity:

dotnet add package Azure.Identity

Instale o pacote semantic-kernel com as dependências opcionais do Azure:

pip install semantic-kernel[azure]

No momento, os agentes não estão disponíveis em Java.

Configurando o cliente de projeto de IA

O acesso a um AzureAIAgent requer primeiro a criação de um cliente de projeto configurado especificamente para um Projeto do Foundry, geralmente por meio do fornecimento de uma cadeia de conexão (O SDK do Azure AI Foundry: Introdução aos Projetos).

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());

O AgentsClient pode ser acessado do AIProjectClient:

AgentsClient agentsClient = client.GetAgentsClient();

Modifique o arquivo .env no diretório raiz para incluir:

AZURE_AI_AGENT_PROJECT_CONNECTION_STRING = "<example-connection-string>"
AZURE_AI_AGENT_MODEL_DEPLOYMENT_NAME = "<example-model-deployment-name>"

ou

AZURE_AI_AGENT_ENDPOINT = "<example-endpoint>"
AZURE_AI_AGENT_SUBSCRIPTION_ID = "<example-subscription-id>"
AZURE_AI_AGENT_RESOURCE_GROUP_NAME = "<example-resource-group-name>"
AZURE_AI_AGENT_PROJECT_NAME = "<example-project-name>"
AZURE_AI_AGENT_MODEL_DEPLOYMENT_NAME = "<example-model-deployment-name>"

Depois que a configuração for definida, o cliente poderá ser criado:

async with (
    DefaultAzureCredential() as creds,
    AzureAIAgent.create_client(credential=creds) as client,
):
    # Your operational code here

No momento, os agentes não estão disponíveis em Java.

Criando um AzureAIAgent

Para criar um AzureAIAgent, comece configurando e inicializando o projeto do agente por meio do serviço de IA do Azure e, em seguida, integre-o ao Kernel Semântico:

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

// 1. Define an agent on the Azure AI agent service
Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>");

// 2. Create a Semantic Kernel agent based on the agent definition
AzureAIAgent agent = new(definition, agentsClient);
from azure.identity.aio import DefaultAzureCredential
from semantic_kernel.agents.azure_ai import AzureAIAgent, AzureAIAgentSettings

ai_agent_settings = AzureAIAgentSettings.create()

async with (
    DefaultAzureCredential() as creds,
    AzureAIAgent.create_client(credential=creds) as client,
):
    # 1. Define an agent on the Azure AI agent service
    agent_definition = await client.agents.create_agent(
        model=ai_agent_settings.model_deployment_name,
        name="<name>",
        instructions="<instructions>",
    )

    # 2. Create a Semantic Kernel agent based on the agent definition
    agent = AzureAIAgent(
        client=client,
        definition=agent_definition,
    )

No momento, os agentes não estão disponíveis em Java.

Interagindo com um AzureAIAgent

A interação com o AzureAIAgent é simples. O agente mantém o histórico de conversas automaticamente usando um thread:

AgentThread thread = await agentsClient.CreateThreadAsync();
try
{
    ChatMessageContent message = new(AuthorRole.User, "<your user input>");
    await agent.AddChatMessageAsync(threadId, message);
    await foreach (ChatMessageContent response in agent.InvokeAsync(thread.Id))
    {
        Console.WriteLine(response.Content);
    }
}
finally
{
    await this.AgentsClient.DeleteThreadAsync(thread.Id);
    await this.AgentsClient.DeleteAgentAsync(agent.Id);
}
USER_INPUTS = ["Hello", "What's your name?"]

thread = await client.agents.create_thread()

try:
    for user_input in USER_INPUTS:
        await agent.add_chat_message(thread_id=thread.id, message=user_input)
        response = await agent.get_response(thread_id=thread.id)
        print(response)
finally:
    await client.agents.delete_thread(thread.id)

Opcionalmente, um agente pode ser invocado como:

for user_input in USER_INPUTS:
    await agent.add_chat_message(thread_id=thread.id, message=user_input)
    async for content in agent.invoke(thread_id=thread.id):
        print(content.content)

Um agente também pode produzir uma resposta transmitida em fluxo:

ChatMessageContent message = new(AuthorRole.User, "<your user input>");
await agent.AddChatMessageAsync(threadId, message);
await foreach (StreamingChatMessageContent response in agent.InvokeStreamingAsync(thread.Id))
{
    Console.Write(response.Content);
}
for user_input in USER_INPUTS:
    await agent.add_chat_message(thread_id=thread.id, message=user_input)
    async for content in agent.invoke_stream(thread_id=thread.id):
        print(content.content, end="", flush=True)

No momento, os agentes não estão disponíveis em Java.

Usando plug-ins com um AzureAIAgent

O Kernel Semântico dá suporte à extensão de um AzureAIAgent com plug-ins personalizados para funcionalidade aprimorada:

Plugin plugin = KernelPluginFactory.CreateFromType<YourPlugin>();
AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>");

AzureAIAgent agent = new(definition, agentsClient, plugins: [plugin]);
from semantic_kernel.functions import kernel_function

class SamplePlugin:
    @kernel_function(description="Provides sample data.")
    def get_data(self) -> str:
        return "Sample data"

ai_agent_settings = AzureAIAgentSettings.create()

async with (
        DefaultAzureCredential() as creds,
        AzureAIAgent.create_client(credential=creds) as client,
    ):
        agent_definition = await client.agents.create_agent(
            model=ai_agent_settings.model_deployment_name,
        )

        agent = AzureAIAgent(
            client=client,
            definition=agent_definition,
            plugins=[SamplePlugin()]
        )

No momento, os agentes não estão disponíveis em Java.

Recursos avançados

Um AzureAIAgent pode aproveitar ferramentas avançadas como:

Interpretador de Código

O Interpretador de Código permite que os agentes escrevam e executem código Python em um ambiente de execução isolado (Interpretador de Código do Serviço de Agente de IA do Azure).

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>",
    tools: [new CodeInterpreterToolDefinition()],
    toolResources:
        new()
        {
            CodeInterpreter = new()
            {
                FileIds = { ... },
            }
        }));

AzureAIAgent agent = new(definition, agentsClient);
from azure.ai.projects.models import CodeInterpreterTool

async with (
        DefaultAzureCredential() as creds,
        AzureAIAgent.create_client(credential=creds) as client,
    ):
        code_interpreter = CodeInterpreterTool()
        agent_definition = await client.agents.create_agent(
            model=ai_agent_settings.model_deployment_name,
            tools=code_interpreter.definitions,
            tool_resources=code_interpreter.resources,
        )

No momento, os agentes não estão disponíveis em Java.

A pesquisa de arquivos oferece aos agentes conhecimentos externos ao seu modelo (Ferramenta de Pesquisa de Arquivos do Serviço de Agente de IA do Azure).

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>",
    tools: [new FileSearchToolDefinition()],
    toolResources:
        new()
        {
            FileSearch = new()
            {
                VectorStoreIds = { ... },
            }
        }));

AzureAIAgent agent = new(definition, agentsClient);
from azure.ai.projects.models import FileSearchTool

async with (
        DefaultAzureCredential() as creds,
        AzureAIAgent.create_client(credential=creds) as client,
    ):
        file_search = FileSearchTool(vector_store_ids=[vector_store.id])
        agent_definition = await client.agents.create_agent(
            model=ai_agent_settings.model_deployment_name,
            tools=file_search.definitions,
            tool_resources=file_search.resources,
        )

No momento, os agentes não estão disponíveis em Java.

Integração do OpenAPI

Conecta seu agente a uma API externa (Como usar o Serviço de Agente de IA do Azure com Ferramentas Especificadas no OpenAPI).

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

string apiJsonSpecification = ...; // An Open API JSON specification

Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>",
    tools: [
        new OpenApiToolDefinition(
            "<api name>", 
            "<api description>", 
            BinaryData.FromString(apiJsonSpecification), 
            new OpenApiAnonymousAuthDetails())
    ],
);

AzureAIAgent agent = new(definition, agentsClient);
from azure.ai.projects.models import OpenApiTool, OpenApiAnonymousAuthDetails

async with (
        DefaultAzureCredential() as creds,
        AzureAIAgent.create_client(credential=creds) as client,
    ):
        openapi_spec_file_path = "sample/filepath/..."
        with open(os.path.join(openapi_spec_file_path, "spec_one.json")) as file_one:
            openapi_spec_one = json.loads(file_one.read())
        with open(os.path.join(openapi_spec_file_path, "spec_two.json")) as file_two:
            openapi_spec_two = json.loads(file_two.read())

        # Note that connection or managed identity auth setup requires additional setup in Azure
        auth = OpenApiAnonymousAuthDetails()
        openapi_tool_one = OpenApiTool(
            name="<name>",
            spec=openapi_spec_one,
            description="<description>",
            auth=auth,
        )
        openapi_tool_two = OpenApiTool(
            name="<name>",
            spec=openapi_spec_two,
            description="<description>",
            auth=auth,
        )

        agent_definition = await client.agents.create_agent(
            model=ai_agent_settings.model_deployment_name,
            tools=openapi_tool_one.definitions + openapi_tool_two.definitions,
        )

No momento, os agentes não estão disponíveis em Java.

Integração do AzureAI Search

Usar um índice existente do Azure AI Search com seu agente (Usar um índice existente do AI Search).

AIProjectClient client = AzureAIAgent.CreateAzureAIClient("<your connection-string>", new AzureCliCredential());
AgentsClient agentsClient = client.GetAgentsClient();

ConnectionsClient cxnClient = client.GetConnectionsClient();
ListConnectionsResponse searchConnections = await cxnClient.GetConnectionsAsync(AzureAIP.ConnectionType.AzureAISearch);
ConnectionResponse searchConnection = searchConnections.Value[0];

Agent definition = agentsClient.CreateAgentAsync(
    "<name of the the model used by the agent>",
    name: "<agent name>",
    description: "<agent description>",
    instructions: "<agent instructions>",
    tools: [new AzureAIP.AzureAISearchToolDefinition()],
    toolResources: new()
    {
        AzureAISearch = new()
        {
            IndexList = { new AzureAIP.IndexResource(searchConnection.Id, "<your index name>") }
        }
    });

AzureAIAgent agent = new(definition, agentsClient);
from azure.ai.projects.models import AzureAISearchTool, ConnectionType

async with (
        DefaultAzureCredential() as creds,
        AzureAIAgent.create_client(credential=creds) as client,
    ):
        conn_list = await client.connections.list()

        ai_search_conn_id = ""
        for conn in conn_list:
            if conn.connection_type == ConnectionType.AZURE_AI_SEARCH:
                ai_search_conn_id = conn.id
                break

        ai_search = AzureAISearchTool(
            index_connection_id=ai_search_conn_id, 
            index_name=AZURE_AI_SEARCH_INDEX_NAME,
        )

        agent_definition = await client.agents.create_agent(
            model=ai_agent_settings.model_deployment_name,
            instructions="Answer questions using your index.",
            tools=ai_search.definitions,
            tool_resources=ai_search.resources,
            headers={"x-ms-enable-preview": "true"},
        )

No momento, os agentes não estão disponíveis em Java.

Recuperando um AzureAIAgent existente

Um agente existente pode ser recuperado e reutilizado ao especificar o ID do assistente:

Agent definition = agentsClient.GetAgentAsync("<your agent id>");
AzureAIAgent agent = new(definition, agentsClient);
agent_definition = await client.agents.get_agent(assistant_id="your-agent-id")
agent = AzureAIAgent(client=client, definition=agent_definition)

No momento, os agentes não estão disponíveis em Java.

Excluindo um AzureAIAgent

Os agentes e seus threads associados podem ser excluídos quando não forem mais necessários:

await agentsClient.DeleteThreadAsync(thread.Id);
await agentsClient.DeleteAgentAsync(agent.Id);
await client.agents.delete_thread(thread.id)
await client.agents.delete_agent(agent.id)

Se estiver trabalhando com um repositório de vetores ou arquivos, eles também poderão ser excluídos:

await agentsClient.DeleteVectorStoreAsync("<your store id>");
await agentsClient.DeleteFileAsync("<your file id>");
await client.agents.delete_file(file_id=file.id)
await client.agents.delete_vector_store(vector_store_id=vector_store.id)

No momento, os agentes não estão disponíveis em Java.

Mais informações sobre a ferramenta de pesquisa de arquivos é descrita no artigo ferramenta de pesquisa de arquivos do Serviço do Agente de IA do Azure artigo.

How-To

Para obter exemplos práticos de como usar um AzureAIAgent, consulte nossos exemplos de código no GitHub:

No momento, os agentes não estão disponíveis em Java.

Colaboração do agente no de chat do agente