Crawl4AI MCP provider

Exposes crawling, Markdown, HTML, screenshot, PDF, JavaScript, and library-context tools to Claude Code over MCP.

Parent repository GitHub

Type: Claude Code integration / tool surface
Repository: unclecode/crawl4ai
Readiness: Usable for development and self-hosted workflows, but deployments should use a current security release and receive careful hardening, resource controls, and legal review.
Keywords: 8

Location

Repository path

deploy/docker/README.md

Invocation

`claude mcp add --transport sse c4ai-sse http://localhost:11235/mcp/sse`

Setup

Installation / activation

Run the Docker server, then register its SSE endpoint in Claude Code; WebSocket is also documented at `/mcp/ws`.

Keywords

Claude CodeMCPtool surfaceSSEWebSocketcrawlscreenshotPDF

Repository context

unclecode/crawl4ai

Apache-2.0 Python web-crawling and scraping framework that produces LLM-ready Markdown and structured data through browser automation, adaptive and deep crawling, CLI, SDK, Docker API, and a documented MCP server that exposes crawl, HTML, screenshot, PDF, JavaScript, and library-context tools directly to Claude Code.

web crawlingweb scrapingbrowser automationdata extractionRAG ingestionagent toolingself-hosted infrastructureweb-to-Markdown

Open full repository research