Yoo guys. I made some websites by now but every time I start making one it feels so AI generated even if i prompt it well. What can i do? I use claude design, i use ui/ux pro max skill inside Claude Code

Day 1 of my 100-build challenge is done. I built an AI Visibility Auditor to check whether a business actually shows up when people ask ChatGPT, Claude, or Perplexity for recommendations. I tested it on a well-established HVAC company in Dallas. They scored 35/100. They looked strong on Google, but AI struggled to clearly understand and describe the business. The auditor: • Tests real customer questions • Checks live AI answers, search signals, and the website • Explains where the score comes from • Gives the top fixes to improve visibility The biggest lesson was simple: Being visible on Google and being understood by AI are not the same thing. I’m sharing the source code for free. Comment “GEO” and I’ll send it over. Day 1 of 100 complete :)

What do you think does Fable 5 will be extended after 19 July too? Want to know your opinion after seeing their extension pattern

I got tired of paying a monthly fee just to talk to my computer, so I built my own and I'm giving it away. It's called Keak. You hold two keys and talk, and it answers you out loud, sees your screen, and actually does things. Say "put it in my calendar for Thursday" and it just does it, without you even looking. It can run routines while you sleep, connect to your Google, Microsoft, Slack and Notion, and hand the big jobs to a team of agents. The part most of you will care about: it runs on your own AI. Sign in with your Claude or ChatGPT and there's no API key. Want Gemini or a free local model with Ollama instead? That works too. Short video above shows what it actually does. If you want to try it, it's at keak.app, free to use. Not trying to sell anyone anything, I just think a lot of you would genuinely use this. Happy to answer whatever in the comments.

The first thing I built was an infographic skill, like Nate in the video. Whaat it does: ✅ Is triggered by natural language or commands. ✅ Is connected via Higgsfield's MCP server and generates images as JPGs using Nano Banana. ✅ Outputs images based on my brand assets. I'm finding it hard to decide what to build next, since I think a lot of skills already exist. Question is, to what extent does it make sense to adapt existing skills and tailor them to my own needs? Not everything will be based on the same framework Nate uses, right? In any case looking forward to the next session ;D #AISChallenge