> mcp // seo, search & web
Web Scraper
for AI.
Web scraping with Cloudflare Browser Rendering, AI-powered data extraction, and R2 storage.
// what it is
Getting clean, usable content out of a web page is harder than it sounds: JavaScript rendering, bot protection, and inconsistent markup get in the way. This server uses Cloudflare Browser Rendering as the primary engine with Firecrawl as a fallback, giving your AI assistant reliable access to page content as markdown, screenshots, PDFs, extracted structured data, or raw link lists, with everything saved to R2 for later retrieval.
GOOD FOR // Developers and AI agents that need reliable page content for research, competitive analysis, content audits, or feeding data into downstream workflows.
// what you can do
Page content in multiple formats
Fetch any page as clean markdown, HTML, or plain text. The browser engine handles JavaScript-rendered pages that simpler scrapers miss.
Screenshots and PDFs
Capture a full-page screenshot or generate a PDF of any URL, stored automatically in R2 and retrievable by your assistant at any time.
AI-powered structured extraction
Describe what you want out of a page (product details, contact information, pricing) and the AI extraction layer returns it as structured JSON rather than raw HTML.
Link extraction and saved pages
Pull all links from a page, list and retrieve previously saved pages, or delete saved content when it is no longer needed.
// try asking
Scrape the services page from this URL and give me the content as clean markdown.
Take a full-page screenshot of the competitor's pricing page and save it.
Extract the product name, price, and description from this e-commerce product page as JSON.
List all the links on the client's homepage.
// the full toolset
scrape Fetch page content (markdown/html/text) Browser binding → Firecrawl fallback json AI-powered structured data extraction REST API `/json` (needs `CF_API_TOKEN`) screenshot Capture page screenshot (PNG → R2) Browser binding (Puppeteer) pdf Generate PDF from page (→ R2) Browser binding (Puppeteer) links Extract all links from page Browser binding (Puppeteer) list List saved pages D1 query get Get saved page content (supports AI summary) R2 read delete Delete a saved page D1 + R2 delete // connect
Works with Claude (Code, Desktop, claude.ai), and any MCP-compatible client. OAuth on first connect: no keys to paste.
claude mcp add --transport http web-scraper-mcp https://scraper.mcp.jezweb.ai/mcpclaude.ai: Settings → Connectors → Add connector → https://scraper.mcp.jezweb.ai/mcp
// more in seo, search & web
Search Console
Search analytics, sitemaps, URL inspection, and multi-account management for agencies with 200+ properties..
Google Analytics
GA4 reporting with custom dimensions, traffic sources, top pages, device breakdowns, and quick reports..
DataForSEO
Comprehensive SEO data from DataForSEO: backlinks, keyword research, SERP analysis, technical audits, AI optimization tracking, business reviews, and more..
Bing Webmaster
Bing Webmaster Tools API for AI agents.
Whispering Wombat
SEO site crawler: crawl sites and analyse technical SEO issues.
Website Content
MCP server providing structured website content for chat agents with visual rendering capabilities..