TypeScript版LlamaIndex来袭！用Agentic RAG构建智能AI代理，效率与隐私双提升！ 原创

# 下载并安装 fnm
winget install Schniz.fnm

# 下载并安装 Node.js
fnm install 22

# 验证 Node.js 版本
node -v # 应输出 "v22.14.0"

# 验证 npm 版本
npm -v # 应输出 "10.9.2"
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.

$ md simple-agent
$ cd simple-agent
$ npm init
$ npm install llamaindex @llamaindex/ollama
1.
2.
3.
4.

import { agent, Settings, tool } from "llamaindex";
import { z } from "zod";
import { Ollama, OllamaEmbedding } from "@llamaindex/ollama";
1.
2.
3.

const llama3 = new Ollama({
  model: "llama3.2:1b",
});
1.
2.
3.

Settings.llm = llama3;
1.

const addNumbers = tool({
  name: "SumNumbers",
  description: "使用此函数计算两个数字的和",
  parameters: z.object({
    a: z.number().describe("第一个数字"),
    b: z.number().describe("第二个数字"),
  }),
  execute: ({ a, b }: { a: number; b: number }) => `${a + b}`,
});
1.
2.
3.
4.
5.
6.
7.
8.
9.

const divideNumbers = tool({
  name: "divideNumber",
  description: "使用此函数计算两个数字的商",
  parameters: z.object({
    a: z.number().describe("被除数"),
    b: z.number().describe("除数"),
  }),
  execute: ({ a, b }: { a: number; b: number }) => `${a / b}`,
});
1.
2.
3.
4.
5.
6.
7.
8.
9.

async function main(query: string) {
const mathAgent = agent({
    tools: [addNumbers, divideNumbers],
    llm: llama3,
    verbose: false,
  });

const response = await mathAgent.run(query);
console.log(response.data);
}

// 驱动代码，运行应用
const query = "Add two number 5 and 7 and divide by 2";
void main(query).then(() => {
console.log("Done");
});
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

const query = "If the total number of boys in a class is 50 and girls is 30, what is the total number of students in the class?";
void main(query).then(() => {
  console.log("Done");
});
1.
2.
3.
4.

$ md agentic-rag-app
$ cd agentic-rag-app
$ npm init
$ npm install llamaindex @llamaindex/ollama
1.
2.
3.
4.

// constant.ts
const constant = {
  STORAGE_DIR: "./storage",
  DATA_FILE: "data/pual-essay.txt",
};

export default constant;
1.
2.
3.
4.
5.
6.
7.

import { Settings, storageContextFromDefaults } from "llamaindex";
import { Ollama, OllamaEmbedding } from "@llamaindex/ollama";
import { Document, VectorStoreIndex } from "llamaindex";
import fs from "fs/promises";
import constant from "./constant";
1.
2.
3.
4.
5.

const llama3 = new Ollama({
  model: "llama3.2:1b",
});

const nomic = new OllamaEmbedding({
  model: "nomic-embed-text",
});
1.
2.
3.
4.
5.
6.
7.

Settings.llm = llama3;
Settings.embedModel = nomic;
1.
2.

async function indexAndStorage() {
try {
    // 设置持久化存储
    const storageContext = await storageContextFromDefaults({
      persistDir: constant.STORAGE_DIR,
    });

    // 加载文档
    const essay = await fs.readFile(constant.DATA_FILE, "utf-8");
    constdocument = new Document({
      text: essay,
      id_: "essay",
    });

    // 创建并持久化索引
    await VectorStoreIndex.fromDocuments([document], {
      storageContext,
    });

    console.log("索引和嵌入文件已成功存储！");
  } catch (error) {
    console.log("索引过程中发生错误：", error);
  }
}
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.

export default indexAndStorage;
1.

import {
  Settings,
  storageContextFromDefaults,
  VectorStoreIndex,
} from "llamaindex";
import constant from "./constant";
import { Ollama, OllamaEmbedding } from "@llamaindex/ollama";
import { agent } from "llamaindex";
1.
2.
3.
4.
5.
6.
7.
8.

async function loadAndQuery(query: string) {
try {
    // 从持久化存储中加载索引
    const storageContext = await storageContextFromDefaults({
      persistDir: constant.STORAGE_DIR,
    });

    // 加载已有的索引
    const index = await VectorStoreIndex.init({ storageContext });

    // 创建检索器和查询引擎
    const retriever = index.asRetriever();
    const queryEngine = index.asQueryEngine({ retriever });

    const tools = [
      index.queryTool({
        metadata: {
          name: "paul_graham_essay_tool",
          description: `此工具可以回答有关 Paul Graham 文章的详细问题。`,
        },
      }),
    ];
    const ragAgent = agent({ tools });

    // 查询已索引的文章
    const response = await queryEngine.query({ query });
    let toolResponse = await ragAgent.run(query);

    console.log("响应：", response.message);
    console.log("工具响应：", toolResponse);
  } catch (error) {
    console.log("检索过程中发生错误：", error);
  }
}
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.

export default loadAndQuery;
1.

import indexAndStorage from"./load-index";
import loadAndQuery from"./query-paul";

function main(query: string) {
console.log("======================================");
console.log("正在索引数据....");
  indexAndStorage();
console.log("数据索引完成！");
console.log("请稍候，正在获取响应或订阅！");
  loadAndQuery(query);
}

const query = "什么是生活？";
main(query);
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

$ npx tsx ./app.ts
1.