Summary

One call, and the anti-bot problem is somebody else's.

The whole surface is a single tool, which is the point: the work that usually breaks a scraper — proxies, bot detection, turning HTML into something a model can read — happens on UseScraper's side. That also means the key you supply carries the cost, and pages fail or succeed according to their service, not your machine.

What it is

A one-tool server in front of the UseScraper API. You give it a URL and it returns the page content in the format you asked for, optionally routed through a proxy that gets past bot detection, and optionally shaped into a structure you define.

What you get

A page's content fetched and returned as markdown by default, or as `text` or `html` — `scrape`
An advanced proxy path for sites that reject ordinary requests, off by default — `advanced_proxy`
Structured extraction: pass `extract_object` to describe the fields you want back instead of the whole page

Requirements

A UseScraper API key in `USESCRAPER_API_KEY`. Runs as a Node process against the built `build/index.js`; also installable into Claude Desktop through Smithery. The package is `usescraper-server` 0.1.0.

Setup effort

One command plus a key — npx -y @smithery/cli install usescraper-server --client claude, then supply credentials