Sql agent

title: 构建自定义 SQL 智能体 sidebarTitle: 自定义 SQL 智能体

在本教程中，我们将使用 LangGraph 构建一个可以回答有关 SQL 数据库问题的自定义智能体。

LangChain 提供了内置的 智能体 实现，这些实现是使用 LangGraph 原语构建的。如果需要更深度的定制，可以直接在 LangGraph 中实现智能体。本指南演示了 SQL 智能体的一个示例实现。如需实践介绍，请参阅 使用更高级别的 LangChain 抽象构建 SQL 智能体。

构建 SQL 数据库的问答系统需要执行模型生成的 SQL 查询。这样做存在固有风险。确保您的数据库连接权限始终尽可能窄地限定在智能体的需求范围内。这将减轻（尽管不能完全消除）构建模型驱动系统的风险。

预构建的智能体 让我们能够快速开始，但我们依赖于系统提示来约束其行为——例如，我们指示智能体始终从“列出表”工具开始，并在执行查询之前始终运行查询检查器工具。

我们可以通过定制智能体在 LangGraph 中实施更高程度的控制。在这里，我们实现了一个简单的 ReAct 智能体设置，具有用于特定工具调用的专用节点。我们将使用与预构建智能体相同的 [状态]。

概念

我们将涵盖以下概念：

• 工具 用于读取 SQL 数据库
• LangGraph 图 API，包括状态、节点、边和条件边。
• 人在回路 流程

设置

安装

pip

pip install langchain  langgraph  langchain-community

LangSmith

设置 LangSmith 以检查链或智能体内部发生的情况。然后设置以下环境变量：

export LANGSMITH_TRACING="true"
export LANGSMITH_API_KEY="..."

1. 选择 LLM

选择一个支持 工具调用 的模型：

OpenAI

👉 阅读 OpenAI 聊天模型集成文档

pip install -U "langchain[openai]"

init_chat_model

import os
from langchain.chat_models import init_chat_model

os.environ["OPENAI_API_KEY"] = "sk-..."

model = init_chat_model("gpt-5.2")

import os
from langchain_openai import ChatOpenAI

os.environ["OPENAI_API_KEY"] = "sk-..."

model = ChatOpenAI(model="gpt-5.2")

Anthropic

👉 阅读 Anthropic 聊天模型集成文档

pip install -U "langchain[anthropic]"

init_chat_model

import os
from langchain.chat_models import init_chat_model

os.environ["ANTHROPIC_API_KEY"] = "sk-..."

model = init_chat_model("claude-sonnet-4-6")

import os
from langchain_anthropic import ChatAnthropic

os.environ["ANTHROPIC_API_KEY"] = "sk-..."

model = ChatAnthropic(model="claude-sonnet-4-6")

Azure

👉 阅读 Azure 聊天模型集成文档

pip install -U "langchain[openai]"

init_chat_model

import os
from langchain.chat_models import init_chat_model

os.environ["AZURE_OPENAI_API_KEY"] = "..."
os.environ["AZURE_OPENAI_ENDPOINT"] = "..."
os.environ["OPENAI_API_VERSION"] = "2025-03-01-preview"

model = init_chat_model(
    "azure_openai:gpt-5.2",
    azure_deployment=os.environ["AZURE_OPENAI_DEPLOYMENT_NAME"],
)

import os
from langchain_openai import AzureChatOpenAI

os.environ["AZURE_OPENAI_API_KEY"] = "..."
os.environ["AZURE_OPENAI_ENDPOINT"] = "..."
os.environ["OPENAI_API_VERSION"] = "2025-03-01-preview"

model = AzureChatOpenAI(
    model="gpt-5.2",
    azure_deployment=os.environ["AZURE_OPENAI_DEPLOYMENT_NAME"]
)

Google Gemini

👉 阅读 Google GenAI 聊天模型集成文档

pip install -U "langchain[google-genai]"

init_chat_model

import os
from langchain.chat_models import init_chat_model

os.environ["GOOGLE_API_KEY"] = "..."

model = init_chat_model("google_genai:gemini-2.5-flash-lite")

import os
from langchain_google_genai import ChatGoogleGenerativeAI

os.environ["GOOGLE_API_KEY"] = "..."

model = ChatGoogleGenerativeAI(model="gemini-2.5-flash-lite")

AWS Bedrock

👉 阅读 AWS Bedrock 聊天模型集成文档

pip install -U "langchain[aws]"

init_chat_model

from langchain.chat_models import init_chat_model

# 按照以下步骤配置您的凭据：
# https://docs.aws.amazon.com/bedrock/latest/userguide/getting-started.html

model = init_chat_model(
    "anthropic.claude-3-5-sonnet-20240620-v1:0",
    model_provider="bedrock_converse",
)

from langchain_aws import ChatBedrock

model = ChatBedrock(model="anthropic.claude-3-5-sonnet-20240620-v1:0")

HuggingFace

👉 阅读 HuggingFace 聊天模型集成文档

pip install -U "langchain[huggingface]"

init_chat_model

import os
from langchain.chat_models import init_chat_model

os.environ["HUGGINGFACEHUB_API_TOKEN"] = "hf_..."

model = init_chat_model(
    "microsoft/Phi-3-mini-4k-instruct",
    model_provider="huggingface",
    temperature=0.7,
    max_tokens=1024,
)

import os
from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint

os.environ["HUGGINGFACEHUB_API_TOKEN"] = "hf_..."

llm = HuggingFaceEndpoint(
    repo_id="microsoft/Phi-3-mini-4k-instruct",
    temperature=0.7,
    max_length=1024,
)
model = ChatHuggingFace(llm=llm)

下面示例中显示的输出使用了 OpenAI。

2. 配置数据库

您将为本教程创建一个 SQLite 数据库。SQLite 是一个轻量级数据库，易于设置和使用。我们将加载 chinook 数据库，这是一个代表数字媒体商店的示例数据库。

为了方便起见，我们将数据库 (Chinook.db) 托管在公共 GCS 存储桶上。

import requests, pathlib

url = "https://storage.googleapis.com/benchmarks-artifacts/chinook/Chinook.db"
local_path = pathlib.Path("Chinook.db")

if local_path.exists():
    print(f"{local_path} already exists, skipping download.")
else:
    response = requests.get(url)
    if response.status_code == 200:
        local_path.write_bytes(response.content)
        print(f"File downloaded and saved as {local_path}")
    else:
        print(f"Failed to download the file. Status code: {response.status_code}")

我们将使用 langchain_community 包中可用的便捷 SQL 数据库包装器来与数据库交互。该包装器提供了一个简单的接口来执行 SQL 查询并获取结果：

from langchain_community.utilities import SQLDatabase

db = SQLDatabase.from_uri("sqlite:///Chinook.db")

print(f"Dialect: {db.dialect}")
print(f"Available tables: {db.get_usable_table_names()}")
print(f'Sample output: {db.run("SELECT * FROM Artist LIMIT 5;")}')

Dialect: sqlite
Available tables: ['Album', 'Artist', 'Customer', 'Employee', 'Genre', 'Invoice', 'InvoiceLine', 'MediaType', 'Playlist', 'PlaylistTrack', 'Track']
Sample output: [(1, 'AC/DC'), (2, 'Accept'), (3, 'Aerosmith'), (4, 'Alanis Morissette'), (5, 'Alice In Chains')]

3. 添加数据库交互工具

使用 langchain_community 包中可用的 SQLDatabase 包装器来与数据库交互。该包装器提供了一个简单的接口来执行 SQL 查询并获取结果：

from langchain_community.agent_toolkits import SQLDatabaseToolkit

toolkit = SQLDatabaseToolkit(db=db, llm=model)

tools = toolkit.get_tools()

for tool in tools:
    print(f"{tool.name}: {tool.description}\n")

sql_db_query: Input to this tool is a detailed and correct SQL query, output is a result from the database. If the query is not correct, an error message will be returned. If an error is returned, rewrite the query, check the query, and try again. If you encounter an issue with Unknown column 'xxxx' in 'field list', use sql_db_schema to query the correct table fields.

sql_db_schema: Input to this tool is a comma-separated list of tables, output is the schema and sample rows for those tables. Be sure that the tables actually exist by calling sql_db_list_tables first! Example Input: table1, table2, table3

sql_db_list_tables: Input is an empty string, output is a comma-separated list of tables in the database.

sql_db_query_checker: Use this tool to double check if your query is correct before executing it. Always use this tool before executing a query with sql_db_query!

4. 定义应用步骤

我们为以下步骤构建专用节点：

• 列出数据库表
• 调用“获取模式”工具
• 生成查询
• 检查查询

将这些步骤放在专用节点中，我们可以 (1) 在需要时强制进行工具调用，以及 (2) 自定义每个步骤关联的提示词。

from typing import Literal

from langchain.messages import AIMessage
from langchain_core.runnables import RunnableConfig
from langgraph.graph import END, START, MessagesState, StateGraph
from langgraph.prebuilt import ToolNode


get_schema_tool = next(tool for tool in tools if tool.name == "sql_db_schema")
get_schema_node = ToolNode([get_schema_tool], name="get_schema")

run_query_tool = next(tool for tool in tools if tool.name == "sql_db_query")
run_query_node = ToolNode([run_query_tool], name="run_query")


# Example: create a predetermined tool call
def list_tables(state: MessagesState):
    tool_call = {
        "name": "sql_db_list_tables",
        "args": {},
        "id": "abc123",
        "type": "tool_call",
    }
    tool_call_message = AIMessage(content="", tool_calls=[tool_call])

    list_tables_tool = next(tool for tool in tools if tool.name == "sql_db_list_tables")
    tool_message = list_tables_tool.invoke(tool_call)
    response = AIMessage(f"Available tables: {tool_message.content}")

    return {"messages": [tool_call_message, tool_message, response]}


# Example: force a model to create a tool call
def call_get_schema(state: MessagesState):
    # Note that LangChain enforces that all models accept `tool_choice="any"`
    # as well as `tool_choice=<string name of tool>`.
    llm_with_tools = model.bind_tools([get_schema_tool], tool_choice="any")
    response = llm_with_tools.invoke(state["messages"])

    return {"messages": [response]}


generate_query_system_prompt = """
You are an agent designed to interact with a SQL database.
Given an input question, create a syntactically correct {dialect} query to run,
then look at the results of the query and return the answer. Unless the user
specifies a specific number of examples they wish to obtain, always limit your
query to at most {top_k} results.

You can order the results by a relevant column to return the most interesting
examples in the database. Never query for all the columns from a specific table,
only ask for the relevant columns given the question.

DO NOT make any DML statements (INSERT, UPDATE, DELETE, DROP etc.) to the database.
""".format(
    dialect=db.dialect,
    top_k=5,
)


def generate_query(state: MessagesState):
    system_message = {
        "role": "system",
        "content": generate_query_system_prompt,
    }
    # We do not force a tool call here, to allow the model to
    # respond naturally when it obtains the solution.
    llm_with_tools = model.bind_tools([run_query_tool])
    response = llm_with_tools.invoke([system_message] + state["messages"])

    return {"messages": [response]}


check_query_system_prompt = """
You are a SQL expert with a strong attention to detail.
Double check the {dialect} query for common mistakes, including:
- Using NOT IN with NULL values
- Using UNION when UNION ALL should have been used
- Using BETWEEN for exclusive ranges
- Data type mismatch in predicates
- Properly quoting identifiers
- Using the correct number of arguments for functions
- Casting to the correct data type
- Using the proper columns for joins

If there are any of the above mistakes, rewrite the query. If there are no mistakes,
just reproduce the original query.

You will call the appropriate tool to execute the query after running this check.
""".format(dialect=db.dialect)


def check_query(state: MessagesState):
    system_message = {
        "role": "system",
        "content": check_query_system_prompt,
    }

    # Generate an artificial user message to check
    tool_call = state["messages"][-1].tool_calls[0]
    user_message = {"role": "user", "content": tool_call["args"]["query"]}
    llm_with_tools = model.bind_tools([run_query_tool], tool_choice="any")
    response = llm_with_tools.invoke([system_message, user_message])
    response.id = state["messages"][-1].id

    return {"messages": [response]}

5. 实现智能体

我们现在可以使用 图 API 将这些步骤组装成工作流。我们在查询生成步骤定义一个 条件边，如果生成了查询，则路由到查询检查器，或者如果没有工具调用（例如 LLM 已提供对查询的响应），则结束。

def should_continue(state: MessagesState) -> Literal[END, "check_query"]:
    messages = state["messages"]
    last_message = messages[-1]
    if not last_message.tool_calls:
        return END
    else:
        return "check_query"


builder = StateGraph(MessagesState)
builder.add_node(list_tables)
builder.add_node(call_get_schema)
builder.add_node(get_schema_node, "get_schema")
builder.add_node(generate_query)
builder.add_node(check_query)
builder.add_node(run_query_node, "run_query")

builder.add_edge(START, "list_tables")
builder.add_edge("list_tables", "call_get_schema")
builder.add_edge("call_get_schema", "get_schema")
builder.add_edge("get_schema", "generate_query")
builder.add_conditional_edges(
    "generate_query",
    should_continue,
)
builder.add_edge("check_query", "run_query")
builder.add_edge("run_query", "generate_query")

agent = builder.compile()

我们在下方可视化应用程序：

from IPython.display import Image, display
from langchain_core.runnables.graph import CurveStyle, MermaidDrawMethod, NodeStyles

display(Image(agent.get_graph().draw_mermaid_png()))

我们现在可以调用图：

question = "Which genre on average has the longest tracks?"

for step in agent.stream(
    {"messages": [{"role": "user", "content": question}]},
    stream_mode="values",
):
    step["messages"][-1].pretty_print()

================================ Human Message =================================

Which genre on average has the longest tracks?
================================== Ai Message ==================================

Available tables: Album, Artist, Customer, Employee, Genre, Invoice, InvoiceLine, MediaType, Playlist, PlaylistTrack, Track
================================== Ai Message ==================================
Tool Calls:
  sql_db_schema (call_yzje0tj7JK3TEzDx4QnRR3lL)
 Call ID: call_yzje0tj7JK3TEzDx4QnRR3lL
  Args:
    table_names: Genre, Track
================================= Tool Message =================================
Name: sql_db_schema


CREATE TABLE "Genre" (
	"GenreId" INTEGER NOT NULL,
	"Name" NVARCHAR(120),
	PRIMARY KEY ("GenreId")
)

/*
3 rows from Genre table:
GenreId	Name
1	Rock
2	Jazz
3	Metal
*/


CREATE TABLE "Track" (
	"TrackId" INTEGER NOT NULL,
	"Name" NVARCHAR(200) NOT NULL,
	"AlbumId" INTEGER,
	"MediaTypeId" INTEGER NOT NULL,
	"GenreId" INTEGER,
	"Composer" NVARCHAR(220),
	"Milliseconds" INTEGER NOT NULL,
	"Bytes" INTEGER,
	"UnitPrice" NUMERIC(10, 2) NOT NULL,
	PRIMARY KEY ("TrackId"),
	FOREIGN KEY("MediaTypeId") REFERENCES "MediaType" ("MediaTypeId"),
	FOREIGN KEY("GenreId") REFERENCES "Genre" ("GenreId"),
	FOREIGN KEY("AlbumId") REFERENCES "Album" ("AlbumId")
)

/*
3 rows from Track table:
TrackId	Name	AlbumId	MediaTypeId	GenreId	Composer	Milliseconds	Bytes	UnitPrice
1	For Those About To Rock (We Salute You)	1	1	1	Angus Young, Malcolm Young, Brian Johnson	343719	11170334	0.99
2	Balls to the Wall	2	2	1	U. Dirkschneider, W. Hoffmann, H. Frank, P. Baltes, S. Kaufmann, G. Hoffmann	342562	5510424	0.99
3	Fast As a Shark	3	2	1	F. Baltes, S. Kaufman, U. Dirkscneider & W. Hoffman	230619	3990994	0.99
*/
================================== Ai Message ==================================
Tool Calls:
  sql_db_query (call_cb9ApLfZLSq7CWg6jd0im90b)
 Call ID: call_cb9ApLfZLSq7CWg6jd0im90b
  Args:
    query: SELECT Genre.Name, AVG(Track.Milliseconds) AS AvgMilliseconds FROM Track JOIN Genre ON Track.GenreId = Genre.GenreId GROUP BY Genre.GenreId ORDER BY AvgMilliseconds DESC LIMIT 5;
================================== Ai Message ==================================
Tool Calls:
  sql_db_query (call_DMVALfnQ4kJsuF3Yl6jxbeAU)
 Call ID: call_DMVALfnQ4kJsuF3Yl6jxbeAU
  Args:
    query: SELECT Genre.Name, AVG(Track.Milliseconds) AS AvgMilliseconds FROM Track JOIN Genre ON Track.GenreId = Genre.GenreId GROUP BY Genre.GenreId ORDER BY AvgMilliseconds DESC LIMIT 5;
================================= Tool Message =================================
Name: sql_db_query

[('Sci Fi & Fantasy', 2911783.0384615385), ('Science Fiction', 2625549.076923077), ('Drama', 2575283.78125), ('TV Shows', 2145041.0215053763), ('Comedy', 1585263.705882353)]
================================== Ai Message ==================================

The genre with the longest tracks on average is "Sci Fi & Fantasy," with an average track length of approximately 2,911,783 milliseconds. Other genres with relatively long tracks include "Science Fiction," "Drama," "TV Shows," and "Comedy."

查看上述运行的 LangSmith 追踪。

6. 实现人在回路审查

在执行 SQL 查询之前检查智能体的查询是否有意外操作或低效之处可能是明智的。

这里我们利用 LangGraph 的 人在回路 功能，在执行 SQL 查询之前暂停运行并等待人工审查。使用 LangGraph 的 持久化层，我们可以无限期地（或至少只要持久化层存活）暂停运行。

让我们将 sql_db_query 工具包装在一个接收人类输入的节点中。我们可以使用 中断 函数来实现这一点。下面，我们允许输入批准工具调用、编辑其参数或提供用户反馈。

from langchain_core.runnables import RunnableConfig
from langchain.tools import tool
from langgraph.types import interrupt

@tool(
    run_query_tool.name,
    description=run_query_tool.description,
    args_schema=run_query_tool.args_schema
)
def run_query_tool_with_interrupt(config: RunnableConfig, **tool_input):
    request = {
        "action": run_query_tool.name,
        "args": tool_input,
        "description": "Please review the tool call"
    }
    response = interrupt([request])
    # approve the tool call
    if response["type"] == "accept":
        tool_response = run_query_tool.invoke(tool_input, config)
    # update tool call args
    elif response["type"] == "edit":
        tool_input = response["args"]["args"]
        tool_response = run_query_tool.invoke(tool_input, config)
    # respond to the LLM with user feedback
    elif response["type"] == "response":
        user_feedback = response["args"]
        tool_response = user_feedback
    else:
        raise ValueError(f"Unsupported interrupt response type: {response['type']}")

    return tool_response

# Redefine the tool node to use the interrupt version
run_query_node = ToolNode([run_query_tool_with_interrupt], name="run_query")

上述实现遵循更广泛的 人在回路 指南中的 工具中断示例。请参考该指南了解详细信息和替代方案。

现在让我们重新组装我们的图。我们将用人工审查替换程序化检查。请注意，我们现在包含一个 检查点；这是暂停和恢复运行所必需的。

from langgraph.checkpoint.memory import InMemorySaver

def should_continue(state: MessagesState) -> Literal[END, "run_query"]:
    messages = state["messages"]
    last_message = messages[-1]
    if not last_message.tool_calls:
        return END
    else:
        return "run_query"

builder = StateGraph(MessagesState)
builder.add_node(list_tables)
builder.add_node(call_get_schema)
builder.add_node(get_schema_node, "get_schema")
builder.add_node(generate_query)
builder.add_node(run_query_node, "run_query")

builder.add_edge(START, "list_tables")
builder.add_edge("list_tables", "call_get_schema")
builder.add_edge("call_get_schema", "get_schema")
builder.add_edge("get_schema", "generate_query")
builder.add_conditional_edges(
    "generate_query",
    should_continue,
)
builder.add_edge("run_query", "generate_query")

checkpointer = InMemorySaver()
agent = builder.compile(checkpointer=checkpointer)

我们可以像以前一样调用图。这次，执行被中断：

import json

config = {"configurable": {"thread_id": "1"}}

question = "Which genre on average has the longest tracks?"

for step in agent.stream(
    {"messages": [{"role": "user", "content": question}]},
    config,
    stream_mode="values",
):
    if "messages" in step:
        step["messages"][-1].pretty_print()
    elif "__interrupt__" in step:
        action = step["__interrupt__"][0]
        print("INTERRUPTED:")
        for request in action.value:
            print(json.dumps(request, indent=2))
    else:
        pass

...

INTERRUPTED:
{
  "action": "sql_db_query",
  "args": {
    "query": "SELECT Genre.Name, AVG(Track.Milliseconds) AS AvgLength FROM Track JOIN Genre ON Track.GenreId = Genre.GenreId GROUP BY Genre.Name ORDER BY AvgLength DESC LIMIT 5;"
  },
  "description": "Please review the tool call"
}

我们可以使用 命令 接受或编辑工具调用：

from langgraph.types import Command


for step in agent.stream(
    Command(resume={"type": "accept"}),
    # Command(resume={"type": "edit", "args": {"query": "..."}}),
    config,
    stream_mode="values",
):
    if "messages" in step:
        step["messages"][-1].pretty_print()
    elif "__interrupt__" in step:
        action = step["__interrupt__"][0]
        print("INTERRUPTED:")
        for request in action.value:
            print(json.dumps(request, indent=2))
    else:
        pass

================================== Ai Message ==================================
Tool Calls:
  sql_db_query (call_t4yXkD6shwdTPuelXEmY3sAY)
 Call ID: call_t4yXkD6shwdTPuelXEmY3sAY
  Args:
    query: SELECT Genre.Name, AVG(Track.Milliseconds) AS AvgLength FROM Track JOIN Genre ON Track.GenreId = Genre.GenreId GROUP BY Genre.Name ORDER BY AvgLength DESC LIMIT 5;
================================= Tool Message =================================
Name: sql_db_query

[('Sci Fi & Fantasy', 2911783.0384615385), ('Science Fiction', 2625549.076923077), ('Drama', 2575283.78125), ('TV Shows', 2145041.0215053763), ('Comedy', 1585263.705882353)]
================================== Ai Message ==================================

The genre with the longest average track length is "Sci Fi & Fantasy" with an average length of about 2,911,783 milliseconds. Other genres with long average track lengths include "Science Fiction," "Drama," "TV Shows," and "Comedy."

有关详细信息，请参阅 人在回路指南。

下一步

查看 评估图 指南，了解如何使用 LangSmith 评估 LangGraph 应用程序，包括此类 SQL 智能体。

---

Edit this page on GitHub or file an issue.

Connect these docs to Claude, VSCode, and more via MCP for real-time answers.