The response rate on 5.1 is 🤯 paired with Atlas.... game changer 👏👏 @OpenAI

New favorite podcast... @ChrissyChaosPod Go check it out.

Friday nights be like- Squashing 🐛🐛🐛's claude code ftw..

Someone once said, “Debugging is storytelling in reverse.” And I don’t know how to feel about it…

The most underrated advantage of building with AI: You can iterate at the speed of curiosity. The ones who make it don’t plan perfectly — they adapt relentlessly. Every failure’s a lesson, not a setback.

Pro tip: Don’t sprint with 25% left in the conversation. Auto-compacts always hit at the worst possible time. Compact first. Sprint after.

Thank you. @vercel You’ve made building focused, seamless, and beautifully integrated. From environments to error logs, integrations to compute — the entire process just works. One of those rare platforms that removes friction instead of adding it. 🙌 #buildinginpublic

I’ve started asking AI, “What would Woz do?” during code sessions. Not joking — it instantly changes how AI approaches problems. Somehow, invoking @stevewoz gets cleaner logic, smarter design, and fewer hallucinations. Simple prompt, big results. #Legend #vibecoding

If you’re uncomfortable, unsure, or overthinking something new — pause and look at where you are today. Has staying comfortable gotten you where you want to be? How you treat yourself is how you’ll treat everything around you — even your code. Push your boundaries and be kind

During AI-assisted code sprints, don’t just rely on the auto-generated to-dos. Create your own .md checklist to track what’s been validated and what’s pending. Why? Because when the chat resets or you hit a rabbit hole, Git history alone won’t tell you where you left off. This

Hitting weekly limits in 4 hours isn't acceptable... I understand that people may have abused some privileges, but power users who are making genuine contributions shouldn't be penalized. Claude/OpenAI/Grok should monitor for abuse, rather than limiting all users based on a

Breaking bad habits starts with learning better ones. Every coding session starts the same way for me: focus on the prompt to achieve better outcomes. Here’s my workflow using the npm package I built npm install -g cuecli cuecli get precision-modification #AI #ClaudeCode

Started building 3 months ago. Now 900 commits deep. 100K+ lines of enterprise-ready code — built with AI. Tomorrow I file the patent. Wish me luck 🙌 Early testers — I’ll need you soon. Who’s in? 🚀

New update from @claudeai dropped. Search history is easier. Directions + actions are clearer. And they fixed the user input blending! Boom! @AnthropicAI #buildinpublic #AI

Updating the favicon is minor. But the moment you see your logo on the site… it suddenly feels real.

Fix 1 bug → spawn 2 more. Know your dependencies before you hit commit.

@claudeai @claudeai bumping this... (btw if it matters, I am a MAX user.. )

If your own thoughts hold you back, you’ll never know if the idea was good. Silence the doubts. Haters will always be there..