Hapi plugin for Prerender.io. Intercepts requests from bots and crawlers and serves prerendered HTML, so your JavaScript-rendered app is fully indexable by search engines and social media scrapers.

Compatible with @hapi/hapi v21+ and Node.js 18+.

npm install prerender-hapi

const Hapi = require('@hapi/hapi');

const server = Hapi.server({ host: 'localhost', port: 3000 });

await server.register({
  plugin: require('prerender-hapi'),
  options: {
    token: 'YOUR_PRERENDER_TOKEN'
  }
});

The plugin registers an onRequest extension that transparently proxies bot requests to Prerender.io and returns the prerendered HTML. Regular browser requests are unaffected.

PRERENDER_TOKEN=your_token_here
PRERENDER_SERVICE_URL=https://service.prerender.io/  # optional

await server.register({
  plugin: require('prerender-hapi'),
  options: {
    serviceUrl: 'http://your-prerender-server:3000'
  }
});

const cache = new Map();

await server.register({
  plugin: require('prerender-hapi'),
  options: {
    token: 'YOUR_PRERENDER_TOKEN',
    beforeRender: async (request) => {
      return cache.get(request.url.href) || null;
    },
    afterRender: (request, response) => {
      cache.set(request.url.href, response);
    }
  }
});

Requests are prerendered when all of the following are true:

• The HTTP method is GET
• The User-Agent matches a known bot/crawler (Googlebot, Bingbot, Twitterbot, GPTBot, ClaudeBot, etc.)
  — OR the URL contains _escaped_fragment_
  — OR the X-Bufferbot header is present
• The URL does not end with a static asset extension (.js, .css, .png, .woff2, .webp, etc.) — matched case-insensitively

Everything else passes through to your normal route handlers.

MIT