@nhinguyen122026/antibot-scraper-client

High-performance Node.js & TypeScript client for the Anti-Bot & TLS Fingerprint Edge Scraper API. Engineered for AI Agents, LLM context retrieval, and modern RAG pipelines.

✨ Features

• 🛡️ Authentic TLS/JA3/JA4 Fingerprint Impersonation: Bypasses Cloudflare, Akamai, and perimeter bot detection by spoofing full browser TLS handshakes.
• 🤖 LLM & Agent Native: Extracts clean, dense Markdown ready to be fed directly into GPT-4o, Claude 3.5 Sonnet, or Gemini 1.5/2.0 context windows.
• ⚡ Zero External Dependencies: Powered by native fetch with AbortController timeout handling. Compatible with Node 18+, Bun, Deno, and Cloudflare Workers.
• 🔄 Autonomous Challenge Retries: Automated fallback retry mechanism with browser profile rotation upon detecting Cloudflare Turnstile or WAF challenges.
• 📘 Full TypeScript Definitions: First-class type safety with detailed JSDoc documentation.

📦 Installation

npm install @nhinguyen122026/antibot-scraper-client
# or
pnpm add @nhinguyen122026/antibot-scraper-client
# or
yarn add @nhinguyen122026/antibot-scraper-client

🚀 Quick Start

1. Basic Markdown Extraction

import { AntiBotScraperClient } from '@nhinguyen122026/antibot-scraper-client';

const scraper = new AntiBotScraperClient({
  apiKey: process.env.RAPIDAPI_KEY // Optional if using free tier
});

async function main() {
  const result = await scraper.scrape({
    url: 'https://news.ycombinator.com',
    output_format: 'markdown',
    profile_id: 'chrome124_win'
  });

  console.log(`Scraped ${result.content_length} bytes in ${result.duration_ms}ms`);
  console.log(`TLS Fingerprint: ${result.tls_fingerprint}`);
  console.log(result.content);
}

main();

2. Streamlined Content Shortcut

const { AntiBotScraperClient } = require('@nhinguyen122026/antibot-scraper-client');

const scraper = new AntiBotScraperClient();

async function run() {
  // Returns raw markdown string directly
  const markdown = await scraper.scrapeContent('https://en.wikipedia.org/wiki/Artificial_intelligence');
  console.log(markdown.slice(0, 500));
}

run();

🤖 RAG & AI Agent Integration

LangChain / LlamaIndex Document Loader Pattern

import { AntiBotScraperClient } from '@nhinguyen122026/antibot-scraper-client';

const scraper = new AntiBotScraperClient();

async function loadWebDocumentForRAG(url: string) {
  const response = await scraper.scrape({
    url,
    output_format: 'markdown',
    timeout: 25
  });

  return {
    pageContent: response.content,
    metadata: {
      source: response.url,
      finalUrl: response.final_url,
      retrievedAt: new Date().toISOString(),
      tlsProfile: response.profile_used
    }
  };
}

⚙️ Configuration & Options

AntiBotScraperClient(options)

• apiKey (string, optional): Your RapidAPI subscription key.
• baseUrl (string, optional): Custom endpoint override.
• defaultTimeout (number, default: 30000): HTTP client timeout in milliseconds.

scraper.scrape(options)

| Option | Type | Default | Description | |---|---|---|---| | url | string | (required) | Target webpage URL to fetch | | output_format | 'markdown' \| 'raw_html' \| 'text' | 'markdown' | Format of extracted content | | profile_id | string | 'chrome124_win' | TLS fingerprint profile: chrome124_win, chrome120_mac, safari17_mac, edge101_win | | timeout | number | 20 | Upstream scraper timeout in seconds | | max_retries | number | 2 | Number of automated challenge retry attempts |

🛡️ Error Handling

import { AntiBotScraperClient, AntiBotScraperError } from '@nhinguyen122026/antibot-scraper-client';

const scraper = new AntiBotScraperClient();

try {
  const res = await scraper.scrape({ url: 'https://protected-site.com' });
} catch (err) {
  if (err instanceof AntiBotScraperError) {
    console.error(`Scrape Error [Status ${err.statusCode}]:`, err.message);
    console.error('Raw details:', err.rawData);
  } else {
    console.error('Network or timeout error:', err);
  }
}

📄 License

MIT © Duong Y Nhi