Scrapling MCP Server

Was er kann

The MCP server that comes with Scrapling, a Python web-scraping library, installed with the scrapling[ai] extra plus browser dependencies and started with the scrapling-mcp command. It runs locally over stdio, or over streamable HTTP on localhost, which refuses to start without a bearer token unless explicitly told otherwise. Its 17 tools make plain HTTP requests, open and reuse browser sessions, save and load browser state such as logins, fetch and extract page content, take screenshots, run page actions and JavaScript, and list the network requests a page made. The README says the server uses Scrapling's stealth browser and Cloudflare solver so the model can get past CAPTCHAs and Cloudflare Turnstile pages, which is a way around sites' bot protection.

Funktionen

Details