Feature request: anybrowse MCP node for agent-native web scraping with Cloudflare bypass

Problem

Flowise workflows that need to scrape web content hit a wall with Cloudflare-protected sites. The HTTP node and most scraping utilities get blocked -- which breaks research agents, content pipelines, and anything that needs real-time web data.

Proposed Solution

A dedicated anybrowse node (or generic MCP tool node that connects to https://anybrowse.dev/mcp).

anybrowse is a web scraping service built specifically for AI agents. It routes requests through residential browser workers, rendering JavaScript and bypassing bot detection before returning clean markdown. It exposes an MCP interface natively.

Available tools via MCP:

• scrape -- fetch any URL, returns markdown + metadata
• search -- web search with structured results
• extract -- LLM-powered structured data extraction
• crawl -- multi-page site crawling
• batch -- parallel multi-URL scraping

Why Flowise is a good fit

Many Flowise use cases involve agents that need to read live web content. An anybrowse node would make those workflows work reliably on real-world sites.

Docs: https://anybrowse.dev/docs npm: anybrowse MCP endpoint: https://anybrowse.dev/mcp