Unbrowse

Every website Unbrowse indexed as its own remote MCP server and agent skill (unofficial). npx skills add unbrowse.ai --skill <site>

Unbrowse - Reverse engineer any website so that your agents can access its APIs directly

Redfin scraper for Node.js: homes for sale in any city or county as structured JSON. Built on Unbrowse.

Google News scraper for Node.js: keyword search as structured JSON, past the 100-result cap. Built on Unbrowse.

Craigslist scraper for Node.js: search results for any city and category as structured JSON. Built on Unbrowse.

Career site jobs scraper for Node.js: every open job on Greenhouse, Lever and Ashby boards as JSON (salary, location, remote). Built on Unbrowse.

YouTube scraper (Node.js): search results and video stats as JSON: views, likes, comments, duration, publish date, tags, channel. No browser, no API quota.

Zillow scraper (Node.js): for-sale, rental and sold listings as JSON: price, beds, baths, sqft, Zestimate, coordinates, photos, with filters. No browser, no proxy pool.

Amazon scraper (Node.js): search results and product pages as JSON: price, rating, reviews, bought-last-month, seller, Best Sellers Rank, specs. No browser, no proxy pool.

open-scrapers: structured scrapers and 1,000+ site APIs for agents and code. Amazon, Zillow and YouTube scrapers in Node.js, plus a catalog of 2,400+ sites as REST/MCP tools.