Does anyone know of a training course step by step that can walk me through how to build websites using Claude Code. Similar to what this guy has done. https://youtu.be/gWNFna6fgS8?si=bgqjC8OzogkHulbi

I've been spending a lot of time the past few weeks to use Claude more and more. After trying Claude Code (a poorly named product because it can do so much more than code) I'm becoming more convinced that everyone beginner needs to understand more use cases and AI products. Claude just happens to be my favorite, and the only way I've been learning about it is because I've been finding resources from other people who are super excited about sharing how they're using things like Claude Code. Curiosity is the key here, and as I've been cruising around for more insights on Claude I realized there isn't a great space online to know you're working with other real humans. I've been having to obsessively piece together portions of tutorials, YouTube videos and online forums. So this is going to be that space for everyone else to have an easier time, have fun, and learn together. If you're curious about AI as a beginner, this community will be all around getting the basics down. It will always be free, forever, and we will curate other communities for folks who are more advanced with AI. If you want to learn from other people about how they're using Claude and other AI tools to have fun and enjoy more of life, this is the community for you.

Hi folks — I’m Vahid Madadi Mahani.I’m very excited about MCP, and I focus on using AI for SEO and digital marketing. I’m exploring Model Context Protocol (MCP) as a way for an AI assistant (e.g., Claude or your model of choice) to interact with n8n workflows and nodes—not just execute tasks, but also validate designs and control the build process. The idea: - Standardize access: AI reads tool docs via MCP, understands inputs/outputs, and selects the right n8n capability. - Design review: Assistant proposes a workflow plan (triggers, nodes, data paths); MCP exposes a validation tool to check schema, auth scopes, and idempotency. - Controlled execution: n8n runs the approved plan with retries, rate limits, and audit logs; AI only sees what MCP allows. Early blueprint I’m considering: 1. Minimal MCP toolbox: search.entities, db.query, cms.publish, queue.enqueue. 2. Guardrails: allowlisted resources, scoped tokens, size limits, content/PII lint. 3. Human-in-the-loop: approvals whenever publishing or touching prod data. 4. Run health: correlate MCP tool calls with n8n execution IDs for end-to-end traceability. For those who’ve tried similar AI-orchestrated builds: - What’s the smallest safe tool surface you’d start with? - Any tips for keeping prompts stable while tools evolve? Appreciate any feedback or gotchas you’ve encountered!

Is anyone using Claude Max Opus 4.1? What is your review?

My main project has been trying to rebuild my website proveworth.com from scratch using Claude Code as a completely nontechnical person. My reasoning is that I want to offer more features like being able to host your own page and have customers be able to submit reviews, like a Yelp but more for communities. I bought the 100$ Claude Max plan and have been chipping away at it every day. It's been frustrating getting the 'You're absolutely right!' as someone who is just barely starting to learn to code, but between switching back and forth between ChatGPT and Claude Code, there has definitely been progress made. So far the biggest unlock has been learning that CC can't actually design well, much less see well. So learning that you can hook up a design based Model Context Protocol inside of the system that can help with the design process feels like a big unlock. I'd love to learn more about what you're experimenting with and what use cases you're finding out in the wild!