Houtini.

What's WebMCP? Take a look at this working* implementation: *I've implemented WebMCP on YubHub. Jobseekers can use YubHub with WebMCP. Developers get a free jobs xml or jobs.json feed. Or, an MCP to build custom job feeds for their sites on top of our data. There's a stats endpoint for all the data too, if you're interested in hiring trends. Post: https://lnkd.in/eaJJ8i-R Website: https://yubhub.co WebMCP inspector: https://lnkd.in/eZ3-Tb-Z

The writing has never been the hard part. Knowing what to write, is. Every morning, it's the same routine. Skim the blogs. Half-watch a YouTube video. Check three subreddits. Try to work out what's actually worth writing about today. Is the content calendar immediately out of date? By the time I had a topic, the morning was gone. And I still hadn't written a word. So I built the thing I wished existed. Content Marketing Ideas reads your industry for you. It monitors the blogs, YouTube channels, subreddits and podcasts you already follow. It reads them every day, transcribes the videos, cross-references everything against your Search Console data, and sends you a daily brief. Not a list of links. A brief. A thesis. Three angles (safe, bold, contrarian). The keywords people are searching. The questions they're actually asking. The gap nobody's covered yet. The part I'm proudest of: if you want a draft article, we can do that with voice matching. 16 statistical analysers read your existing, published work and it measures how you write. Sentence rhythm. How you open. How direct you are. When the tool drafts an article, it matches your patterns, not a template. You still do the final pass, but you're editing something that already sounds like you. When you're happy, click a button and it pushes to your WordPress drafts. https://lnkd.in/egYbVxWg

MCP protocol is probably more powerful than you realise. Welcome, SEO Crawler MCP - a production-grade SEO crawler / auditing tool that demonstrates the full power of Anthropic's Model Context Protocol (MCP) SDK. What you're seeing here is a complete application running inside an MCP container! MCP can host sophisticated, multi-step workflows with their own databases, crawlers, and analysis engines. We've barely scratched the surface of what is possible in the MCP wrapper - this isn't really an MCP *server*, more of an MCP Application. At its core, SEO Crawler MCP uses: Crawlee HttpCrawler - A production-ready web scraping framework from Apify inspired by the Librecrawl extraction engine (ported to Node from Python) that handles: - Smart request scheduling and queue management - Automatic retry logic and error handling - Respectful crawling with configurable delays - Memory-efficient streaming for large sites - Better-SQLite3 - An embedded SQL database that stores: - Every crawled page's HTML, metadata, and headers - Link relationships and site structure - Pre-computed metrics for fast analysis - Custom SQL Analysis Engine - 25+ specialized queries that detect: - Content issues (duplicate titles, missing H1s, thin content) - Technical SEO problems (broken links, canonical issues) - Security vulnerabilities (missing headers, unsafe links) - Optimisation opportunities (title length, meta descriptions) Claude can orchestrate this entire stack through simple function calls (you can also run it in terminal for crawls over 1000 pages) The crawl runs asynchronously, stores everything in SQLite, and then Claude can query that data through natural language - "analyse this crawl for seo opportunities" or "report on internal broken links" - and the MCP server translates that into sophisticated SQL analysis. In this video I'm crawling Houtini.com, which has a few problems! Broken Internal Links Duplicate Title Tags (26 pages) Missing H1 Tags (4 pages) Security Headers (100 pages) - The entire site is missing HSTS headers. Missing Meta Descriptions (100 pages) Thin Content (5 pages) I'm going to deploy this as soon as I cover off a few small issues (I'm currently fixing a small meta header detection issue. If you'd like to test it out, let me know in the comments. Thanks!