Scrapling MCP Server

by Karim Shoair (D4Vinci)

Added to Onei Oct 3, 2026

1 views

v0.4.15

BSD-3-Clause

Repository Docs

Web & Search

Summary

Lets AI agents scrape the web through Scrapling, with HTTP, browser and Cloudflare-bypassing stealth fetches and CSS-narrowed, sanitized output.

Features

13 tools: one-shot and bulk HTTP requests, browser fetches and stealth fetches
Stealth mode built to bypass anti-bot systems such as Cloudflare Turnstile
Persistent browser and request sessions for multi-step browsing
CSS selector narrowing to cut token usage
Prompt-injection content stripped from pages before the model sees them
Screenshot tool and remote browser support over CDP
Token-authenticated HTTP transport in addition to stdio

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

Open your agent's MCP configuration file.
Merge the snippet below into it.
Restart the agent — the "Scrapling MCP Server" tools will be available.

{
  "mcpServers": {
    "scrapling-mcp": {
      "command": "scrapling-mcp"
    }
  }
}

Claude Code

Add the server with a single CLI command.

claude mcp add scrapling-mcp -- scrapling-mcp

Claude Desktop

Add the server to claude_desktop_config.json (Settings → Developer → Edit Config).

{
  "mcpServers": {
    "scrapling-mcp": {
      "command": "scrapling-mcp"
    }
  }
}

HTTP transport with bearer-token auth

{
  "command": "scrapling-mcp",
  "args": [
    "--http",
    "--auth-token",
    "<your-secret-token>"
  ]
}

Description

Scrapling MCP Server is the built-in Model Context Protocol server of Scrapling, the open-source Python web scraping framework by Karim Shoair. It gives Claude, Cursor and other MCP clients 13 tools for fetching pages: one-shot and bulk plain HTTP requests (any method), full browser fetches, and stealth fetches designed to get past anti-bot systems such as Cloudflare Turnstile. Persistent sessions can be opened, listed and closed so an agent can browse across several steps, and a screenshot tool captures what the browser sees.

The server is built to keep agents cheap and safe: pages can be narrowed with CSS selectors before they reach the model, so it reads less text and spends fewer tokens, and content that looks like prompt injection (such as hidden text) is stripped out first. It also supports remote browsers over CDP and an HTTP transport that is token-authenticated by default.

Install it with pip install "scrapling[ai]", run scrapling install once to fetch browser dependencies, then point your client at the scrapling-mcp executable. Scrapling requires Python 3.10 or newer and is released under the BSD-3-Clause licence. A good fit for research agents, price and content monitoring, and any workflow where plain fetch tools get blocked by modern sites.

Related MCP Servers

MCP: X (Twitter)

by X

Featured

X's own hosted MCP server: full-archive post search, user and timeline lookup, trends, bookmarks and article drafting, under your account's permissions.

Web & Search

MCP: Meilisearch MCP Server

by Meilisearch

Featured

Official MCP server for Meilisearch — create and configure indexes, add documents, run hybrid and semantic searches, and manage API keys and tasks through conversation.

Web & Search Databases & Data

MCP: Apify Actors

by Apify

Featured

Apify's official MCP server — turn thousands of ready-made scrapers into tools your assistant can call.

Web & Search

MCP: Browserbase

by Browserbase

Featured

Browserbase's official MCP server — cloud browser automation with Stagehand's natural-language actions.

Web & Search

Browse all MCP servers →