Scrapling MCP Server

Name: Scrapling MCP Server
Author: D4Vinci

@D4Vinci

リポジトリホームページドキュメント

基本情報

カテゴリ

ブラウザ自動化

ライセンス

BSD-3-Clause

トランスポート

stdio, streamable-http

認証方式

none

作者

D4Vinci

Scrapling MCP Server について

Official MCP server of the Scrapling web scraping framework: HTTP requests, stealthy browser, screenshots and network inspection for AI agents.

#web-scraping #browser-automation #python#stealth-browser#crawling #markdown

Scrapling MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

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

Scrapling MCP Server のツール

Fast HTTP request with any method and browser fingerprint impersonation (matching TLS, headers, HTTP/3).

Render JavaScript and scrape a dynamic page once with the stealthy browser, including Cloudflare Turnstile/Interstitial handling.

Open a persistent stealthy browser session reused by browser_fetch calls.

Open a persistent HTTP session (no browser) that keeps cookies, connections and fingerprint.

Close a browser or HTTP session and free its resources.

List active sessions and their effective settings.

Save cookies, local storage and IndexedDB to a Playwright-compatible JSON file to reuse a login.

Restore a saved state file into an open browser session.

Render and scrape a URL through an open browser session, optionally returning a structured page snapshot with element references.

HTTP request through a session opened with open_request_session.

Capture the viewport, full page or one element as PNG/JPEG without reloading.

Return a short sequence of images to show motion and changes on the page.

Read the current page or an element as an AI snapshot, HTML, Markdown or text, with regex search.

Chain clicks, typing, keyboard shortcuts, dialog replies, scrolling and waits in one call.

Run JavaScript on the current page and return the result as JSON.

List completed network requests across tabs and page loads to find the APIs behind a page.

Inspect one recorded request: summary, headers, sent data or response body.

概要

What Scrapling MCP does

Scrapling is an open-source Python web scraping framework by D4Vinci (BSD-3-Clause). Its built-in MCP server gives Claude, Cursor and other MCP clients 17 tools for HTTP requests, a stealthy browser, screenshots, saved login state, browser actions and network-history inspection.

Install

Scrapling needs Python 3.10 or later.

pip install "scrapling[ai]"   # MCP server dependencies
scrapling install             # browser dependencies (not needed for plain HTTP requests)

A Docker image with all extras and browsers is also published: docker pull pyd4vinci/scrapling (or ghcr.io/d4vinci/scrapling:latest).

Add it to your client

Claude Desktop / Cursor / Windsurf (claude_desktop_config.json):

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

If the client cannot find the command, use the full path from which scrapling-mcp (macOS) or where scrapling-mcp (Windows). With Docker, use "command": "docker", "args": ["run", "-i", "--rm", "pyd4vinci/scrapling", "mcp"].

Claude Code:

claude mcp add ScraplingServer "/path/to/.venv/bin/scrapling-mcp"

The scrapling-mcp command was added in v0.4.13; on older versions run scrapling mcp.

Remote (Streamable HTTP) mode

scrapling-mcp --http serves the tools over Streamable HTTP and refuses to start without authentication. Pass a token, for example scrapling-mcp --http --auth-token "$(openssl rand -hex 32)", or set SCRAPLING_MCP_AUTH_TOKEN in Docker. By default it only listens on 127.0.0.1; --host 0.0.0.0 exposes it to the network. --no-auth exists for local testing only.

Limits and safety

Anyone who can reach an HTTP-mode server can fetch any URL from that machine, which is why authentication is required.
Browser tools block about 3,500 ad and tracker domains and strip hidden content (hidden elements, zero-width characters, HTML comments) that could carry prompt injection.
Respect each site's terms of service and robots rules; bypassing anti-bot protection may not be allowed everywhere.

Sources: Scrapling MCP Server guide, GitHub. Checked 2026-11-07.