Skip to main content
createDeepAgent 具有以下配置选项:
const agent = createDeepAgent({
  model?: BaseLanguageModel | string,
  tools?: TTools | StructuredTool[],
  systemPrompt?: string | SystemMessage,
  middleware?: TMiddleware,
  subagents?: TSubagents,
  responseFormat?: TResponse,
  backend?: AnyBackendProtocol | ((config) => AnyBackendProtocol),
  interruptOn?: Record<string, boolean | InterruptOnConfig>,
  memory?: string[],
  skills?: string[],
  ...
});
有关完整参数列表,请参阅 createDeepAgent API 参考。

模型

provider:model 格式传递 model 字符串,或传递已初始化的模型实例。默认为 anthropic:claude-sonnet-4-6。请参阅支持的模型了解所有提供商,以及推荐模型了解经过测试的推荐。
使用 provider:model 格式(例如 openai:gpt-5)可以快速在模型之间切换。
👉 阅读 OpenAI 聊天模型集成文档
npm install @langchain/openai deepagents
import { createDeepAgent } from "deepagents";

process.env.OPENAI_API_KEY = "your-api-key";

const agent = createDeepAgent({ model: "gpt-5.4" });
// 这将使用默认参数为指定模型调用 initChatModel
// 要使用特定模型参数,请直接使用 initChatModel

连接弹性

LangChain 聊天模型会自动重试失败的 API 请求,并采用指数退避策略。默认情况下,对于网络错误、速率限制(429)和服务器错误(5xx),模型最多重试 6 次。客户端错误(如 401 未授权或 404)不会重试。 您可以在创建模型时调整 maxRetries 参数,以根据您的环境调整此行为:
import { ChatAnthropic } from "@langchain/anthropic";
import { createDeepAgent } from "deepagents";

const agent = createDeepAgent({
    model: new ChatAnthropic({
        model: "claude-sonnet-4-6",
        maxRetries: 10, // 对于不可靠网络增加重试次数(默认:6)
        timeout: 120_000, // 对于慢速连接增加超时时间
    }),
});
对于在不可靠网络上运行的长时间代理任务,考虑将 max_retries 增加到 10-15,并与检查点配对,以便在失败时保留进度。

工具

除了用于规划、文件管理和子代理生成的内置工具外,您还可以提供自定义工具:
import { tool } from "langchain";
import { TavilySearch } from "@langchain/tavily";
import { createDeepAgent } from "deepagents";
import { z } from "zod";

const internetSearch = tool(
  async ({
    query,
    maxResults = 5,
    topic = "general",
    includeRawContent = false,
  }: {
    query: string;
    maxResults?: number;
    topic?: "general" | "news" | "finance";
    includeRawContent?: boolean;
  }) => {
    const tavilySearch = new TavilySearch({
      maxResults,
      tavilyApiKey: process.env.TAVILY_API_KEY,
      includeRawContent,
      topic,
    });
    return await tavilySearch._call({ query });
  },
  {
    name: "internet_search",
    description: "运行网络搜索",
    schema: z.object({
      query: z.string().describe("搜索查询"),
      maxResults: z.number().optional().default(5),
      topic: z
        .enum(["general", "news", "finance"])
        .optional()
        .default("general"),
      includeRawContent: z.boolean().optional().default(false),
    }),
  },
);

const agent = createDeepAgent({
  tools: [internetSearch],
});

系统提示

深度代理附带内置系统提示。默认系统提示包含使用内置规划工具、文件系统工具和子代理的详细说明。 当中间件添加特殊工具(如文件系统工具)时,它们会附加到系统提示中。 每个深度代理还应包含一个针对其特定用例的自定义系统提示:
import { createDeepAgent } from "deepagents";

const researchInstructions = `您是一位专家研究员。` +
  `您的工作是进行彻底研究,然后` +
  `撰写一份精炼的报告。`;

const agent = createDeepAgent({
  systemPrompt: researchInstructions,
});

中间件

默认情况下,深度代理可以访问以下中间件 如果您使用记忆、技能或人机回环,则还包括以下中间件:
  • MemoryMiddleware:当提供 memory 参数时,跨会话持久化和检索对话上下文
  • SkillsMiddleware:当提供 skills 参数时启用自定义技能
  • HumanInTheLoopMiddleware:当提供 interruptOn 参数时,在指定点暂停以等待人工批准或输入

预构建中间件

LangChain 公开了额外的预构建中间件,允许您添加各种功能,例如重试、回退或 PII 检测。有关更多信息,请参阅预构建中间件 deepagents 包还公开了 createSummarizationMiddleware 用于相同的工作流程。有关详细信息,请参阅摘要

特定于提供商的中间件

有关针对特定 LLM 提供商优化的特定于提供商的中间件,请参阅官方集成社区集成

自定义中间件

您可以提供额外的中间件来扩展功能、添加工具或实现自定义钩子:
import { tool, createMiddleware } from "langchain";
import { createDeepAgent } from "deepagents";
import * as z from "zod";

const getWeather = tool(
  ({ city }: { city: string }) => {
    return `The weather in ${city} is sunny.`;
  },
  {
    name: "get_weather",
    description: "获取城市的天气。",
    schema: z.object({
      city: z.string(),
    }),
  }
);

let callCount = 0;

const logToolCallsMiddleware = createMiddleware({
  name: "LogToolCallsMiddleware",
  wrapToolCall: async (request, handler) => {
    // 拦截并记录每个工具调用 - 演示横切关注点
    callCount += 1;
    const toolName = request.toolCall.name;

    console.log(`[Middleware] Tool call #${callCount}: ${toolName}`);
    console.log(
      `[Middleware] Arguments: ${JSON.stringify(request.toolCall.args)}`
    );

    // 执行工具调用
    const result = await handler(request);

    // 记录结果
    console.log(`[Middleware] Tool call #${callCount} completed`);

    return result;
  },
});

const agent = await createDeepAgent({
  model: "google_genai:gemini-3.1-pro-preview",
  tools: [getWeather] as any,
  middleware: [logToolCallsMiddleware] as any,
});
初始化后不要修改属性如果需要在钩子调用之间跟踪值(例如计数器或累积数据),请使用图状态。 图状态按设计作用于线程,因此在并发下更新是安全的。这样做:
const customMiddleware = createMiddleware({
  name: "CustomMiddleware",
  beforeAgent: async (state) => {
    return { x: (state.x ?? 0) + 1 }; // 改为更新图状态
  },
});
不要这样做:
let x = 1;

const customMiddleware = createMiddleware({
  name: "CustomMiddleware",
  beforeAgent: async () => {
    x += 1; // 修改会导致竞态条件
  },
});
原地修改,例如在 beforeAgent 中修改 state.x、在 beforeAgent 中修改共享变量,或在钩子中更改其他共享值,可能会导致细微的错误和竞态条件,因为许多操作是并发运行的(子代理、并行工具和不同线程上的并行调用)。有关使用自定义属性扩展状态的完整详细信息,请参阅自定义中间件 - 自定义状态模式。 如果必须在自定义中间件中使用修改,请考虑当子代理、并行工具或并发代理调用同时运行时会发生什么。

子代理

为了隔离详细工作并避免上下文膨胀,请使用子代理:
import { tool } from "langchain";
import { TavilySearch } from "@langchain/tavily";
import { createDeepAgent, type SubAgent } from "deepagents";
import { z } from "zod";

const internetSearch = tool(
  async ({
    query,
    maxResults = 5,
    topic = "general",
    includeRawContent = false,
  }: {
    query: string;
    maxResults?: number;
    topic?: "general" | "news" | "finance";
    includeRawContent?: boolean;
  }) => {
    const tavilySearch = new TavilySearch({
      maxResults,
      tavilyApiKey: process.env.TAVILY_API_KEY,
      includeRawContent,
      topic,
    });
    return await tavilySearch._call({ query });
  },
  {
    name: "internet_search",
    description: "Run a web search",
    schema: z.object({
      query: z.string().describe("The search query"),
      maxResults: z.number().optional().default(5),
      topic: z
        .enum(["general", "news", "finance"])
        .optional()
        .default("general"),
      includeRawContent: z.boolean().optional().default(false),
    }),
  },
);

const researchSubagent: SubAgent = {
  name: "research-agent",
  description: "Used to research more in depth questions",
  systemPrompt: "You are a great researcher",
  tools: [internetSearch],
  model: "openai:gpt-5.2",  // Optional override, defaults to main agent model
};
const subagents = [researchSubagent];

const agent = createDeepAgent({
  model: "claude-sonnet-4-6",
  subagents,
});
有关更多信息,请参阅子代理

后端

深度代理工具可以利用虚拟文件系统来存储、访问和编辑文件。默认情况下,深度代理使用 StateBackend 如果您使用技能记忆,则必须在创建代理之前将预期的技能或记忆文件添加到后端。
存储在 langgraph 状态中的临时文件系统后端。此文件系统仅持久化单个线程
import { createDeepAgent, StateBackend } from "deepagents";

// 默认情况下,我们提供一个 StateBackend
const agent = createDeepAgent();

// 在底层,它看起来像这样
const agent2 = createDeepAgent({
  backend: new StateBackend(),
});
有关更多信息,请参阅后端

沙盒

沙盒是专门的后端,在隔离环境中运行代理代码,具有自己的文件系统和用于 shell 命令的 execute 工具。 当您希望深度代理写入文件、安装依赖项并运行命令而不更改本地机器上的任何内容时,请使用沙盒后端。 您可以通过在创建深度代理时将沙盒后端传递给 backend 来配置沙盒:
import { createDeepAgent } from "deepagents";
import { ChatAnthropic } from "@langchain/anthropic";
import { DenoSandbox } from "@langchain/deno";

// 创建并初始化沙箱
const sandbox = await DenoSandbox.create({
  memoryMb: 1024,
  lifetime: "10m",
});

try {
  const agent = createDeepAgent({
    model: new ChatAnthropic({ model: "claude-opus-4-6" }),
    systemPrompt: "You are a JavaScript coding assistant with sandbox access.",
    backend: sandbox,
  });

  const result = await agent.invoke({
    messages: [
      {
        role: "user",
        content:
          "Create a simple HTTP server using Deno.serve and test it with curl",
      },
    ],
  });
} finally {
  await sandbox.close();
}
有关更多信息,请参阅沙盒

人机回环

某些工具操作可能很敏感,需要在执行前获得人工批准。 您可以为每个工具配置批准:
import { tool } from "langchain";
import { createDeepAgent } from "deepagents";
import { MemorySaver } from "@langchain/langgraph";
import { z } from "zod";

const deleteFile = tool(
  async ({ path }: { path: string }) => {
    return `Deleted ${path}`;
  },
  {
    name: "delete_file",
    description: "Delete a file from the filesystem.",
    schema: z.object({
      path: z.string(),
    }),
  },
);

const readFile = tool(
  async ({ path }: { path: string }) => {
    return `Contents of ${path}`;
  },
  {
    name: "read_file",
    description: "Read a file from the filesystem.",
    schema: z.object({
      path: z.string(),
    }),
  },
);

const sendEmail = tool(
  async ({ to, subject, body }: { to: string; subject: string; body: string }) => {
    return `Sent email to ${to}`;
  },
  {
    name: "send_email",
    description: "Send an email.",
    schema: z.object({
      to: z.string(),
      subject: z.string(),
      body: z.string(),
    }),
  },
);

// Checkpointer is REQUIRED for human-in-the-loop
const checkpointer = new MemorySaver();

const agent = createDeepAgent({
  model: "google_genai:gemini-3.1-pro-preview",
  tools: [deleteFile, readFile, sendEmail],
  interruptOn: {
    delete_file: true,  // Default: approve, edit, reject
    read_file: false,   // No interrupts needed
    send_email: { allowedDecisions: ["approve", "reject"] },  // No editing
  },
  checkpointer,  // Required!
});
您可以为代理和子代理配置在工具调用时以及从工具调用内部中断。 有关更多信息,请参阅人机回环

技能

您可以使用技能为您的深度代理提供新的功能和专业知识。 虽然工具倾向于涵盖较低级别的功能(如本机文件系统操作或规划),但技能可以包含有关如何完成任务、参考信息和其他资产(如模板)的详细说明。 这些文件仅在代理确定该技能对当前提示有用时才由代理加载。 这种渐进式披露减少了代理在启动时必须考虑的令牌和上下文数量。 有关技能示例,请参阅深度代理示例技能 要将技能添加到您的深度代理,请将它们作为参数传递给 create_deep_agent
import { createDeepAgent, type FileData } from "deepagents";
import { MemorySaver } from "@langchain/langgraph";

const checkpointer = new MemorySaver();

function createFileData(content: string): FileData {
  const now = new Date().toISOString();
  return {
    content: content.split("\n"),
    created_at: now,
    modified_at: now,
  };
}

const skillsFiles: Record<string, FileData> = {};

const skillUrl =
  "https://raw.githubusercontent.com/langchain-ai/deepagentsjs/refs/heads/main/examples/skills/langgraph-docs/SKILL.md";
const response = await fetch(skillUrl);
const skillContent = await response.text();

skillsFiles["/skills/langgraph-docs/SKILL.md"] = createFileData(skillContent);

const agent = await createDeepAgent({
  checkpointer,
  // IMPORTANT: deepagents skill source paths are virtual (POSIX) paths relative to the backend root.
  skills: ["/skills/"],
});

const config = {
  configurable: {
    thread_id: `thread-${Date.now()}`,
  },
};

const result = await agent.invoke(
  {
    messages: [
      {
        role: "user",
        content: "what is langraph? Use the langgraph-docs skill if available.",
      },
    ],
    files: skillsFiles,
  },
  config,
);

记忆

使用 AGENTS.md 文件 为您的深度代理提供额外的上下文。 您可以在创建深度代理时将一个或多个文件路径传递给 memory 参数:
import { createDeepAgent, type FileData } from "deepagents";
import { MemorySaver } from "@langchain/langgraph";

const AGENTS_MD_URL =
  "https://raw.githubusercontent.com/langchain-ai/deepagents/refs/heads/main/examples/text-to-sql-agent/AGENTS.md";

async function fetchText(url: string): Promise<string> {
  const res = await fetch(url);
  if (!res.ok) {
    throw new Error(`Failed to fetch ${url}: ${res.status} ${res.statusText}`);
  }
  return await res.text();
}

const agentsMd = await fetchText(AGENTS_MD_URL);
const checkpointer = new MemorySaver();

function createFileData(content: string): FileData {
  const now = new Date().toISOString();
  return {
    content,
    mimeType: "text/plain",
    created_at: now,
    modified_at: now,
  };
}

const agent = await createDeepAgent({
  memory: ["/AGENTS.md"],
  checkpointer: checkpointer,
});

const result = await agent.invoke(
  {
    messages: [
      {
        role: "user",
        content: "请告诉我您的记忆文件中有什么内容。",
      },
    ],
    // 为默认 StateBackend 的内部状态文件系统播种(虚拟路径必须以 "/" 开头)。
    files: { "/AGENTS.md": createFileData(agentsMd) },
  },
  { configurable: { thread_id: "12345" } }
);

结构化输出

深度代理支持结构化输出 您可以通过将所需结构化输出模式作为 responseFormat 参数传递给 createDeepAgent() 调用来设置它。 当模型生成结构化数据时,它会被捕获、验证,并在代理状态的 ‘structuredResponse’ 键中返回。
import { tool } from "langchain";
import { TavilySearch } from "@langchain/tavily";
import { createDeepAgent } from "deepagents";
import { z } from "zod";

const internetSearch = tool(
  async ({
    query,
    maxResults = 5,
    topic = "general",
    includeRawContent = false,
  }: {
    query: string;
    maxResults?: number;
    topic?: "general" | "news" | "finance";
    includeRawContent?: boolean;
  }) => {
    const tavilySearch = new TavilySearch({
      maxResults,
      tavilyApiKey: process.env.TAVILY_API_KEY,
      includeRawContent,
      topic,
    });
    return await tavilySearch._call({ query });
  },
  {
    name: "internet_search",
    description: "运行网络搜索",
    schema: z.object({
      query: z.string().describe("搜索查询"),
      maxResults: z.number().optional().default(5),
      topic: z
        .enum(["general", "news", "finance"])
        .optional()
        .default("general"),
      includeRawContent: z.boolean().optional().default(false),
    }),
  }
);

const weatherReportSchema = z.object({
  location: z.string().describe("此天气报告的位置"),
  temperature: z.number().describe("当前温度(摄氏度)"),
  condition: z
    .string()
    .describe("当前天气状况(例如,晴朗、多云、下雨)"),
  humidity: z.number().describe("湿度百分比"),
  windSpeed: z.number().describe("风速(公里/小时)"),
  forecast: z.string().describe("未来 24 小时的简要预报"),
});

const agent = await createDeepAgent({
  responseFormat: weatherReportSchema,
  tools: [internetSearch],
});

const result = await agent.invoke({
  messages: [
    {
      role: "user",
      content: "旧金山的天气怎么样?",
    },
  ],
});

console.log(result.structuredResponse);
// {
//   location: 'San Francisco, California',
//   temperature: 18.3,
//   condition: 'Sunny',
//   humidity: 48,
//   windSpeed: 7.6,
//   forecast: 'Clear skies with temperatures remaining mild. High of 18°C (64°F) during the day, dropping to around 11°C (52°F) at night.'
// }
有关更多信息和示例,请参阅响应格式。 有关更多信息和示例,请参阅响应格式