展示HN:我构建了一个SDK,它可以扰乱HTML,让爬虫获取到无用信息。
Show HN: I built an SDK that scrambles HTML so scrapers get garbage

原始链接: https://www.obscrd.dev/

## 使用 obscrd 保护您的内容 obscrd 是一个开源的内容保护系统,旨在保护网站免受抓取和人工智能机器人侵害。它采用多层方法,混淆 HTML 并阻止恶意爬虫,同时不影响用户体验。 **主要功能包括:** * **客户端保护:** 混淆文本,保护电子邮件/电话号码,保护图像(使用 canvas 渲染),并拦截剪贴板以防止轻松复制。 * **服务器端阻止:** 自动执行 `robots.txt` 以阻止超过 20 种人工智能爬虫,并为流行的框架(如 Express 和 Fastify)提供中间件支持。 * **轻松集成:** 利用 React 组件 (`@obscrd/react`) 和 CLI 工具,实现快速的项目设置和种子生成。 obscrd 对人类正常渲染内容,同时向抓取器呈现混淆数据,确保用户获得视觉一致性,并为您的宝贵内容提供强大的保护。加入候补名单以获取抢先体验和更新!

一位独立开发者创建了“obscrd”,一个免费且开源的SDK(可在GitHub上找到:https://github.com/obscrd/obscrd),旨在阻止AI网络爬虫。该工具通过使用CSS视觉上重新排列HTML内容,同时保持用户可读性,从而有效地向依赖纯文本提取的爬虫提供乱码。 除了HTML混淆之外,obscrd还包括电子邮件/电话号码混淆、AI蜜罐(向爬虫注入指令)、剪贴板拦截以及取证面包屑来检测盗窃等功能。它还通过`robots.txt`阻止超过30种已知的AI爬虫。 开发者承认存在局限性——它无法阻止使用无头浏览器或OCR的坚定攻击者——但旨在*提高*简单机器人爬取的成本。Hacker News评论中的一个关键讨论点集中在使用体验问题上,特别是高亮显示和复制粘贴功能损坏,以及对AI生成文档有效性的质疑。开发者澄清说,通过仔细的CSS实现和ARIA属性,为屏幕阅读器的可访问性*是*被保留的。
相关文章

原文
Skip to main content

Protect your contentfrom scrapers and AI bots

Open-source content protection system for React. Scramble your HTML, block AI crawlers, and defend your content at every layer.

app.tsx
import { ObscrdProvider, ProtectedText } from '@obscrd/react'

export default function App() {
  return (
    
      
        This text is readable by humans
        but scrambled for scrapers.
      
    
  )
}

Your content is being scraped right now. obscrd stops it.

See it in action

obscrd uses CSS ordering and decoy character injection. Users see normal text — scrapers reading textContent get garbled nonsense.

What users see

obscrd scrambles your HTML while CSS reconstructs it visually. Scrapers and AI bots only see meaningless characters — real users see perfectly normal text.

Contact us at [email protected]

The obscrd protection stack

Together, they make scraping your site expensive at every layer.

@obscrd/reactClient-side protection
  • – Text obfuscation — CSS flex ordering + character shuffle
  • – Email & phone protection — RTL reversal + decoys
  • – Image protection — canvas rendering, no URLs
  • – Clipboard interception — copy produces shuffled text
  • – AI honeypots & forensic breadcrumbs
npm i @obscrd/react
@obscrd/robotsServer-side crawler blocking
  • – robots.txt automation — block 20+ AI crawlers
  • – Middleware support — Express, Fastify, Node.js
  • – Meta tag generation for crawler control
  • – Works standalone — no React dependency
npm i @obscrd/robots
obscrd CLIProject setup & tooling
  • – Generate cryptographic seeds
  • – Auto-detects .env.local / .env
  • – One command project initialization
npx obscrd init
$ npm install @obscrd/react @obscrd/robots$ npx obscrd init
import { ObscrdProvider, ProtectedText, ProtectedEmail } from '@obscrd/react'

function App() {
  return (
    
      
        This text is readable by humans but scrambled for scrapers.
      
      
    
  )
}

Answers to your questions

obscrd renders your content normally for real users through the browser, but transforms the underlying DOM so that scrapers, bots, and automated tools only see scrambled, meaningless data. The visual output is identical — the source code is not.

Get early access

We're actively building obscrd. Join the waitlist to get updates, early access to new features, and help shape the roadmap.

We respect your inbox. No spam, just product updates. Privacy Policy.

联系我们 contact @ memedata.com