Ovaj vodič pokazuje kako integrirati Model Context Protocol (MCP) servere s Azure AI Foundry agentima, omogućujući snažnu orkestraciju alata i AI mogućnosti za poduzeća.
Model Context Protocol (MCP) je otvoreni standard koji omogućuje AI aplikacijama sigurno povezivanje s vanjskim izvorima podataka i alatima. Kada se integrira s Azure AI Foundry, MCP omogućuje agentima pristup i interakciju s različitim vanjskim uslugama, API-jima i izvorima podataka na standardiziran način.
Ova integracija spaja fleksibilnost MCP-ovog ekosustava alata s robusnim okvirom Azure AI Foundry agenata, pružajući AI rješenja razine poduzeća s opsežnim mogućnostima prilagodbe.
Note: Ako želite koristiti MCP u Azure AI Foundry Agent Service, trenutno su podržane samo sljedeće regije: westus, westus2, uaenorth, southindia i switzerlandnorth
Na kraju ovog vodiča moći ćete:
- Razumjeti Model Context Protocol i njegove prednosti
- Postaviti MCP servere za korištenje s Azure AI Foundry agentima
- Kreirati i konfigurirati agente s MCP integracijom alata
- Implementirati praktične primjere koristeći stvarne MCP servere
- Upravljati odgovorima alata i citatima u razgovorima agenata
Prije početka, provjerite imate li:
- Azure pretplatu s pristupom AI Foundry
- Python 3.10+ ili .NET 8.0+
- Instaliran i konfiguriran Azure CLI
- Odgovarajuće dozvole za kreiranje AI resursa
Model Context Protocol je standardizirani način za AI aplikacije da se povežu s vanjskim izvorima podataka i alatima. Ključne prednosti uključuju:
- Standardizirana integracija: Dosljedno sučelje za različite alate i usluge
- Sigurnost: Sigurni mehanizmi autentifikacije i autorizacije
- Fleksibilnost: Podrška za razne izvore podataka, API-je i prilagođene alate
- Proširivost: Jednostavno dodavanje novih mogućnosti i integracija
Odaberite željeno razvojno okruženje:
Note Možete pokrenuti ovaj notebook
pip install azure-ai-projects -U
pip install azure-ai-agents==1.1.0b4 -U
pip install azure-identity -U
pip install mcp==1.11.0 -Uimport os, time
from azure.ai.projects import AIProjectClient
from azure.identity import DefaultAzureCredential
from azure.ai.agents.models import McpTool, RequiredMcpToolCall, SubmitToolApprovalAction, ToolApprovalmcp_server_url = os.environ.get("MCP_SERVER_URL", "https://learn.microsoft.com/api/mcp")
mcp_server_label = os.environ.get("MCP_SERVER_LABEL", "mslearn")project_client = AIProjectClient(
endpoint="https://your-project-endpoint.services.ai.azure.com/api/projects/your-project",
credential=DefaultAzureCredential(),
)mcp_tool = McpTool(
server_label=mcp_server_label,
server_url=mcp_server_url,
allowed_tools=[], # Optional: specify allowed tools
)with project_client:
agents_client = project_client.agents
# Create a new agent with MCP tools
agent = agents_client.create_agent(
model="Your AOAI Model Deployment",
name="my-mcp-agent",
instructions="You are a helpful agent that can use MCP tools to assist users. Use the available MCP tools to answer questions and perform tasks.",
tools=mcp_tool.definitions,
)
print(f"Created agent, ID: {agent.id}")
print(f"MCP Server: {mcp_tool.server_label} at {mcp_tool.server_url}")
# Create thread for communication
thread = agents_client.threads.create()
print(f"Created thread, ID: {thread.id}")
# Create message to thread
message = agents_client.messages.create(
thread_id=thread.id,
role="user",
content="What's difference between Azure OpenAI and OpenAI?",
)
print(f"Created message, ID: {message.id}")
# Handle tool approvals and run agent
mcp_tool.update_headers("SuperSecret", "123456")
run = agents_client.runs.create(thread_id=thread.id, agent_id=agent.id, tool_resources=mcp_tool.resources)
print(f"Created run, ID: {run.id}")
while run.status in ["queued", "in_progress", "requires_action"]:
time.sleep(1)
run = agents_client.runs.get(thread_id=thread.id, run_id=run.id)
if run.status == "requires_action" and isinstance(run.required_action, SubmitToolApprovalAction):
tool_calls = run.required_action.submit_tool_approval.tool_calls
if not tool_calls:
print("No tool calls provided - cancelling run")
agents_client.runs.cancel(thread_id=thread.id, run_id=run.id)
break
tool_approvals = []
for tool_call in tool_calls:
if isinstance(tool_call, RequiredMcpToolCall):
try:
print(f"Approving tool call: {tool_call}")
tool_approvals.append(
ToolApproval(
tool_call_id=tool_call.id,
approve=True,
headers=mcp_tool.headers,
)
)
except Exception as e:
print(f"Error approving tool_call {tool_call.id}: {e}")
if tool_approvals:
agents_client.runs.submit_tool_outputs(
thread_id=thread.id, run_id=run.id, tool_approvals=tool_approvals
)
print(f"Current run status: {run.status}")
print(f"Run completed with status: {run.status}")
# Display conversation
messages = agents_client.messages.list(thread_id=thread.id)
print("\nConversation:")
print("-" * 50)
for msg in messages:
if msg.text_messages:
last_text = msg.text_messages[-1]
print(f"{msg.role.upper()}: {last_text.text.value}")
print("-" * 50)Note Možete pokrenuti ovaj notebook
#r "nuget: Azure.AI.Agents.Persistent, 1.1.0-beta.4"
#r "nuget: Azure.Identity, 1.14.2"using Azure.AI.Agents.Persistent;
using Azure.Identity;var projectEndpoint = "https://your-project-endpoint.services.ai.azure.com/api/projects/your-project";
var modelDeploymentName = "Your AOAI Model Deployment";
var mcpServerUrl = "https://learn.microsoft.com/api/mcp";
var mcpServerLabel = "mslearn";
PersistentAgentsClient agentClient = new(projectEndpoint, new DefaultAzureCredential());MCPToolDefinition mcpTool = new(mcpServerLabel, mcpServerUrl);PersistentAgent agent = await agentClient.Administration.CreateAgentAsync(
model: modelDeploymentName,
name: "my-learn-agent",
instructions: "You are a helpful agent that can use MCP tools to assist users. Use the available MCP tools to answer questions and perform tasks.",
tools: [mcpTool]
);// Create thread and message
PersistentAgentThread thread = await agentClient.Threads.CreateThreadAsync();
PersistentThreadMessage message = await agentClient.Messages.CreateMessageAsync(
thread.Id,
MessageRole.User,
"What's difference between Azure OpenAI and OpenAI?");
// Configure tool resources with headers
MCPToolResource mcpToolResource = new(mcpServerLabel);
mcpToolResource.UpdateHeader("SuperSecret", "123456");
ToolResources toolResources = mcpToolResource.ToToolResources();
// Create and handle run
ThreadRun run = await agentClient.Runs.CreateRunAsync(thread, agent, toolResources);
while (run.Status == RunStatus.Queued || run.Status == RunStatus.InProgress || run.Status == RunStatus.RequiresAction)
{
await Task.Delay(TimeSpan.FromMilliseconds(1000));
run = await agentClient.Runs.GetRunAsync(thread.Id, run.Id);
if (run.Status == RunStatus.RequiresAction && run.RequiredAction is SubmitToolApprovalAction toolApprovalAction)
{
var toolApprovals = new List<ToolApproval>();
foreach (var toolCall in toolApprovalAction.SubmitToolApproval.ToolCalls)
{
if (toolCall is RequiredMcpToolCall mcpToolCall)
{
Console.WriteLine($"Approving MCP tool call: {mcpToolCall.Name}");
toolApprovals.Add(new ToolApproval(mcpToolCall.Id, approve: true)
{
Headers = { ["SuperSecret"] = "123456" }
});
}
}
if (toolApprovals.Count > 0)
{
run = await agentClient.Runs.SubmitToolOutputsToRunAsync(thread.Id, run.Id, toolApprovals: toolApprovals);
}
}
}
// Display messages
using Azure;
AsyncPageable<PersistentThreadMessage> messages = agentClient.Messages.GetMessagesAsync(
threadId: thread.Id,
order: ListSortOrder.Ascending
);
await foreach (PersistentThreadMessage threadMessage in messages)
{
Console.Write($"{threadMessage.CreatedAt:yyyy-MM-dd HH:mm:ss} - {threadMessage.Role,10}: ");
foreach (MessageContent contentItem in threadMessage.ContentItems)
{
if (contentItem is MessageTextContent textItem)
{
Console.Write(textItem.Text);
}
else if (contentItem is MessageImageFileContent imageFileItem)
{
Console.Write($"<image from ID: {imageFileItem.FileId}>");
}
Console.WriteLine();
}
}Prilikom konfiguriranja MCP alata za vašeg agenta, možete specificirati nekoliko važnih parametara:
mcp_tool = McpTool(
server_label="unique_server_name", # Identifier for the MCP server
server_url="https://api.example.com/mcp", # MCP server endpoint
allowed_tools=[], # Optional: specify allowed tools
)MCPToolDefinition mcpTool = new(
"unique_server_name", // Server label
"https://api.example.com/mcp" // MCP server URL
);Obje implementacije podržavaju prilagođena zaglavlja za autentifikaciju:
mcp_tool.update_headers("SuperSecret", "123456")MCPToolResource mcpToolResource = new(mcpServerLabel);
mcpToolResource.UpdateHeader("SuperSecret", "123456");- Provjerite je li URL MCP servera dostupan
- Provjerite vjerodajnice za autentifikaciju
- Osigurajte mrežnu povezanost
- Pregledajte argumente i formatiranje poziva alata
- Provjerite zahtjeve specifične za server
- Implementirajte ispravno rukovanje greškama
- Optimizirajte učestalost poziva alata
- Koristite keširanje gdje je prikladno
- Pratite vrijeme odziva servera
Za dodatno unapređenje vaše MCP integracije:
- Istražite prilagođene MCP servere: Izgradite vlastite MCP servere za vlasničke izvore podataka
- Implementirajte naprednu sigurnost: Dodajte OAuth2 ili prilagođene mehanizme autentifikacije
- Praćenje i analitika: Implementirajte zapisivanje i nadzor korištenja alata
- Škala rješenja: Razmotrite balansiranje opterećenja i distribuirane arhitekture MCP servera
- Azure AI Foundry dokumentacija
- Primjeri Model Context Protocol
- Pregled Azure AI Foundry agenata
- MCP specifikacija
Za dodatnu podršku i pitanja:
- Pregledajte Azure AI Foundry dokumentaciju
- Provjerite MCP zajedničke resurse
Odricanje od odgovornosti:
Ovaj dokument je preveden korištenjem AI usluge za prevođenje Co-op Translator. Iako težimo točnosti, imajte na umu da automatski prijevodi mogu sadržavati pogreške ili netočnosti. Izvorni dokument na izvornom jeziku treba smatrati autoritativnim izvorom. Za kritične informacije preporučuje se profesionalni ljudski prijevod. Ne snosimo odgovornost za bilo kakva nesporazuma ili pogrešna tumačenja koja proizlaze iz korištenja ovog prijevoda.