Condividi tramite

Strumento per l'interprete di codice Python

Azure Databricks fornisce system.ai.python_exec, una funzione predefinita del catalogo unity che consente agli agenti di intelligenza artificiale di eseguire dinamicamente Python codice scritto dall'agente, fornito da un utente o recuperato da una codebase. È disponibile per impostazione predefinita e può essere usato direttamente in una query SQL:

SELECT python_exec("""
import random
numbers = [random.random() for _ in range(10)]
print(numbers)
""")

Per altre informazioni sugli strumenti dell'agente, vedere strumenti dell'agente di intelligenza artificiale.

Aggiungere l'interprete di codice all'agente

Per aggiungere python_exec all'agente, connettersi al server MCP gestito per lo schema Unity Catalog. L'interprete di codice è disponibile come strumento MCP preconfigurato all'indirizzo https://<workspace-hostname>/api/2.0/mcp/functions/system/ai/python_exec.

OpenAI Agents SDK (App)

from agents import Agent, Runner
from databricks.sdk import WorkspaceClient
from databricks_openai.agents import McpServer

# WorkspaceClient picks up credentials from the environment (Databricks Apps, notebook, CLI)
workspace_client = WorkspaceClient()
host = workspace_client.config.host

# The context manager manages the MCP connection lifecycle and ensures cleanup on exit.
# from_uc_function constructs the endpoint URL from UC identifiers and wires in auth
# from workspace_client, avoiding hardcoded URLs and manual token handling.
async with McpServer.from_uc_function(
    catalog="system",
    schema="ai",
    function_name="python_exec",
    workspace_client=workspace_client,
    name="code-interpreter",
) as code_interpreter:
    agent = Agent(
        name="Coding agent",
        instructions="You are a helpful coding assistant. Use the python_exec tool to run code.",
        model="databricks-claude-sonnet-4-5",
        mcp_servers=[code_interpreter],
    )
    result = await Runner.run(agent, "Calculate the first 10 Fibonacci numbers")
    print(result.final_output)

Concedere all'app l'accesso alla funzione in databricks.yml:

resources:
  apps:
    my_agent_app:
      resources:
        - name: 'python_exec'
          uc_securable:
            securable_full_name: 'system.ai.python_exec'
            securable_type: 'FUNCTION'
            permission: 'EXECUTE'

LangGraph (App)

from databricks.sdk import WorkspaceClient
from databricks_langchain import ChatDatabricks, DatabricksMCPServer, DatabricksMultiServerMCPClient
from langgraph.prebuilt import create_react_agent

workspace_client = WorkspaceClient()
host = workspace_client.config.host

# DatabricksMultiServerMCPClient provides a unified get_tools() interface across
# multiple MCP servers, making it easy to add more servers later without refactoring.
mcp_client = DatabricksMultiServerMCPClient([
    DatabricksMCPServer(
        name="code-interpreter",
        url=f"{host}/api/2.0/mcp/functions/system/ai/python_exec",
        workspace_client=workspace_client,
    ),
])

async with mcp_client:
    tools = await mcp_client.get_tools()
    agent = create_react_agent(
        ChatDatabricks(endpoint="databricks-claude-sonnet-4-5"),
        tools=tools,
    )
    result = await agent.ainvoke(
        {"messages": [{"role": "user", "content": "Calculate the first 10 Fibonacci numbers"}]}
    )
    # LangGraph returns the full conversation history; the last message is the agent's final response
    print(result["messages"][-1].content)

Concedere all'app l'accesso alla funzione in databricks.yml:

resources:
  apps:
    my_agent_app:
      resources:
        - name: 'python_exec'
          uc_securable:
            securable_full_name: 'system.ai.python_exec'
            securable_type: 'FUNCTION'
            permission: 'EXECUTE'

Gestione dei modelli

from databricks.sdk import WorkspaceClient
from databricks_mcp import DatabricksMCPClient
import mlflow

workspace_client = WorkspaceClient()
host = workspace_client.config.host

mcp_client = DatabricksMCPClient(
    server_url=f"{host}/api/2.0/mcp/functions/system/ai/python_exec",
    workspace_client=workspace_client,
)

tools = mcp_client.list_tools()

# get_databricks_resources() extracts the UC permissions the agent needs at runtime.
# Passing these to log_model lets Model Serving grant access automatically at deployment,
# without requiring manual permission configuration.
mlflow.pyfunc.log_model(
    "agent",
    python_model=my_agent,
    resources=mcp_client.get_databricks_resources(),
)

Per distribuire l'agente, vedere Distribuire un agente per le applicazioni di intelligenza artificiale generative (Model Serving). Per informazioni dettagliate sulla registrazione degli agenti con risorse MCP, vedere Usare server MCP gestiti di Databricks.

Passaggi successivi

  • Last updated on