使用 Vercel AI SDK 进行追踪（仅限 JS/TS）

你可以使用 LangSmith 来追踪 Vercel AI SDK 的运行。本指南将通过一个示例进行讲解。

安装

此包装器需要 AI SDK v5 和 langsmith>=0.3.63。如果你使用的是旧版本的 AI SDK 或 langsmith，请参阅基于 OpenTelemetry (OTEL) 的方法，在此页面。

安装 Vercel AI SDK。本指南在下面的代码片段中使用 Vercel 的 OpenAI 集成，但你也可以使用他们的任何其他选项。

npmyarnpnpm

bash

npm install ai @ai-sdk/openai zod

bash

yarn add ai @ai-sdk/openai zod

bash

pnpm add ai @ai-sdk/openai zod

环境配置

Shell

bash

export LANGSMITH_TRACING=true
export LANGSMITH_API_KEY=<你的-api-key>

# 示例使用 OpenAI，但你可以使用任何选择的 LLM 提供商
export OPENAI_API_KEY=<你的-openai-api-key>

# 对于链接到多个工作区的 LangSmith API 密钥，设置 LANGSMITH_WORKSPACE_ID 环境变量以指定要使用的工作区。
export LANGSMITH_WORKSPACE_ID=<你的-workspace-id>

基本设置

导入并包装 AI SDK 方法，然后像往常一样使用它们：

typescript

import { openai } from "@ai-sdk/openai";
import * as ai from "ai";

import { wrapAISDK } from "langsmith/experimental/vercel";

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai);

await generateText({
  model: openai("gpt-5-nano"),
  prompt: "Write a vegetarian lasagna recipe for 4 people.",
});

你应该能在 LangSmith 仪表板中看到一个追踪记录，类似这样。

你也可以追踪带有工具调用的运行：

typescript

import * as ai from "ai";
import { tool, stepCountIs } from "ai";
import { openai } from "@ai-sdk/openai";
import { z } from "zod";

import { wrapAISDK } from "langsmith/experimental/vercel";

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai);

await generateText({
  model: openai("gpt-5-nano"),
  messages: [
    {
      role: "user",
      content: "What are my orders and where are they? My user ID is 123",
    },
  ],
  tools: {
    listOrders: tool({
      description: "list all orders",
      inputSchema: z.object({ userId: z.string() }),
      execute: async ({ userId }) =>
        `User ${userId} has the following orders: 1`,
    }),
    viewTrackingInformation: tool({
      description: "view tracking information for a specific order",
      inputSchema: z.object({ orderId: z.string() }),
      execute: async ({ orderId }) =>
        `Here is the tracking information for ${orderId}`,
    }),
  },
  stopWhen: stepCountIs(5),
});

这将产生一个追踪记录，类似这样。

你可以像往常一样使用其他 AI SDK 方法。

使用 `traceable`

你可以在 AI SDK 调用周围或 AI SDK 工具调用内部包装 traceable 调用。如果你想在 LangSmith 中将运行分组在一起，这很有用：

typescript

import * as ai from "ai";
import { tool, stepCountIs } from "ai";
import { openai } from "@ai-sdk/openai";
import { z } from "zod";

import { traceable } from "langsmith/traceable";
import { wrapAISDK } from "langsmith/experimental/vercel";

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai);

const wrapper = traceable(async (input: string) => {
  const { text } = await generateText({
    model: openai("gpt-5-nano"),
    messages: [
      {
        role: "user",
        content: input,
      },
    ],
    tools: {
      listOrders: tool({
        description: "list all orders",
        inputSchema: z.object({ userId: z.string() }),
        execute: async ({ userId }) =>
          `User ${userId} has the following orders: 1`,
      }),
      viewTrackingInformation: tool({
        description: "view tracking information for a specific order",
        inputSchema: z.object({ orderId: z.string() }),
        execute: async ({ orderId }) =>
          `Here is the tracking information for ${orderId}`,
      }),
    },
    stopWhen: stepCountIs(5),
  });
  return text;
}, {
  name: "wrapper",
});

await wrapper("What are my orders and where are they? My user ID is 123.");

生成的追踪记录将看起来像这样。

在无服务器环境中追踪

在无服务器环境中追踪时，必须在环境关闭前等待所有运行刷新完毕。为此，你可以在包装 AI SDK 方法时传递一个 LangSmith Client 实例，然后调用 await client.awaitPendingTraceBatches()。确保也将其传递给你创建的任何 traceable 包装器：

typescript

import * as ai from "ai";
import { tool, stepCountIs } from "ai";
import { openai } from "@ai-sdk/openai";
import { z } from "zod";

import { Client } from "langsmith";
import { traceable } from "langsmith/traceable";
import { wrapAISDK } from "langsmith/experimental/vercel";

const client = new Client();

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai, { client });

const wrapper = traceable(async (input: string) => {
  const { text } = await generateText({
    model: openai("gpt-5-nano"),
    messages: [
      {
        role: "user",
        content: input,
      },
    ],
    tools: {
      listOrders: tool({
        description: "list all orders",
        inputSchema: z.object({ userId: z.string() }),
        execute: async ({ userId }) =>
          `User ${userId} has the following orders: 1`,
      }),
      viewTrackingInformation: tool({
        description: "view tracking information for a specific order",
        inputSchema: z.object({ orderId: z.string() }),
        execute: async ({ orderId }) =>
          `Here is the tracking information for ${orderId}`,
      }),
    },
    stopWhen: stepCountIs(5),
  });
  return text;
}, {
  name: "wrapper",
  client,
});

try {
  await wrapper("What are my orders and where are they? My user ID is 123.");
} finally {
  await client.awaitPendingTraceBatches();
}

如果你使用 Next.js，有一个方便的 after 钩子，你可以在其中放置此逻辑：

typescript

import { after } from "next/server"
import { Client } from "langsmith";

export async function POST(request: Request) {
  const client = new Client();

  ...

  after(async () => {
    await client.awaitPendingTraceBatches();
  });

  return new Response(JSON.stringify({ ... }), {
    status: 200,
    headers: { "Content-Type": "application/json" },
  });
};

有关更多详细信息，包括在无服务器环境中管理速率限制的信息，请参阅此页面。

传递 LangSmith 配置

你可以在最初包装 AI SDK 方法时以及通过 providerOptions.langsmith 运行它们时，将 LangSmith 特定的配置传递给包装器。这包括元数据（稍后可用于在 LangSmith 中过滤运行）、顶级运行名称、标签、自定义客户端实例等。

包装时传递的配置将应用于你使用包装方法进行的所有未来调用：

typescript

import { openai } from "@ai-sdk/openai";
import * as ai from "ai";

import { wrapAISDK } from "langsmith/experimental/vercel";

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai, {
    metadata: {
      key_for_all_runs: "value",
    },
    tags: ["myrun"],
  });

await generateText({
  model: openai("gpt-5-nano"),
  prompt: "Write a vegetarian lasagna recipe for 4 people.",
});

而通过 providerOptions.langsmith 在运行时传递的配置仅适用于该次运行。我们建议导入并使用 createLangSmithProviderOptions 包装你的配置，以确保正确的类型：

typescript

import { openai } from "@ai-sdk/openai";
import * as ai from "ai";

import {
  wrapAISDK,
  createLangSmithProviderOptions,
} from "langsmith/experimental/vercel";

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai);

const lsConfig = createLangSmithProviderOptions({
  metadata: {
    individual_key: "value",
  },
  name: "my_individual_run",
});

await generateText({
  model: openai("gpt-5-nano"),
  prompt: "Write a vegetarian lasagna recipe for 4 people.",
  providerOptions: {
    langsmith: lsConfig,
  },
});

数据脱敏

你可以通过指定自定义的输入/输出处理函数，来定制 AI SDK 发送到 LangSmith 的输入和输出。如果你正在处理敏感数据并希望避免将其发送到 LangSmith，这很有用。

由于输出格式因你使用的 AI SDK 方法而异，我们建议单独定义配置并将其传递到包装的方法中。你还需要为 AI SDK 调用内部的子 LLM 运行提供单独的函数，因为在顶层调用 generateText 会在内部调用 LLM，并且可能多次调用。

我们还建议向 createLangSmithProviderOptions 传递一个泛型参数，以获取输入和输出的正确类型。以下是 generateText 的示例：

typescript

import {
  wrapAISDK,
  createLangSmithProviderOptions,
} from "langsmith/experimental/vercel";
import * as ai from "ai";
import { openai } from "@ai-sdk/openai";

const { generateText } = wrapAISDK(ai);

const lsConfig = createLangSmithProviderOptions<typeof generateText>({
  processInputs: (inputs) => {
    const { messages } = inputs;
    return {
      messages: messages?.map((message) => ({
        providerMetadata: message.providerOptions,
        role: "assistant",
        content: "REDACTED",
      })),
      prompt: "REDACTED",
    };
  },
  processOutputs: (outputs) => {
    return {
      providerMetadata: outputs.providerMetadata,
      role: "assistant",
      content: "REDACTED",
    };
  },
  processChildLLMRunInputs: (inputs) => {
    const { prompt } = inputs;
    return {
      messages: prompt.map((message) => ({
        ...message,
        content: "REDACTED CHILD INPUTS",
      })),
    };
  },
  processChildLLMRunOutputs: (outputs) => {
    return {
      providerMetadata: outputs.providerMetadata,
      content: "REDACTED CHILD OUTPUTS",
      role: "assistant",
    };
  },
});

const { text } = await generateText({
  model: openai("gpt-5-nano"),
  prompt: "What is the capital of France?",
  providerOptions: {
    langsmith: lsConfig,
  },
});

// Paris.
console.log(text);

实际的返回值将包含原始的、未脱敏的结果，但 LangSmith 中的追踪记录将被脱敏。这是一个示例。

要对工具输入/输出进行脱敏，请像这样将你的 execute 方法包装在 traceable 中：

typescript

import * as ai from "ai";
import { tool, stepCountIs } from "ai";
import { openai } from "@ai-sdk/openai";
import { z } from "zod";

import { Client } from "langsmith";
import { traceable } from "langsmith/traceable";
import { wrapAISDK } from "langsmith/experimental/vercel";

const client = new Client();

const { generateText, streamText, generateObject, streamObject } =
  wrapAISDK(ai, { client });

const { text } = await generateText({
  model: openai("gpt-5-nano"),
  messages: [
    {
      role: "user",
      content: "What are my orders? My user ID is 123.",
    },
  ],
  tools: {
    listOrders: tool({
      description: "list all orders",
      inputSchema: z.object({ userId: z.string() }),
      execute: traceable(
        async ({ userId }) => {
          return `User ${userId} has the following orders: 1`;
        },
        {
          processInputs: (input) => ({ text: "REDACTED" }),
          processOutputs: (outputs) => ({ text: "REDACTED" }),
          run_type: "tool",
          name: "listOrders",
        }
      ) as (input: { userId: string }) => Promise<string>,
    }),
  },
  stopWhen: stepCountIs(5),
});

traceable 的返回类型很复杂，这使得类型转换成为必要。如果你希望避免类型转换，也可以省略 AI SDK 的 tool 包装函数。

使用 Vercel AI SDK 进行追踪（仅限 JS/TS）

安装 ​

环境配置 ​

基本设置 ​

使用 traceable ​

在无服务器环境中追踪 ​

传递 LangSmith 配置 ​

数据脱敏 ​

安装

环境配置

基本设置

使用 `traceable`

在无服务器环境中追踪

传递 LangSmith 配置

数据脱敏