@thsottiaux exportable project (conversations and contexts) so that another machine (or user) could start "magically" from the same exact point without needing to explaining all from scratch
@trq212 I think people are smart enough to understand the difference between personal and commercial (where it seems obvious that API must be used). The question is simple: for "personal use" can we use our Claude Code (in my case MAX) subscription for openclaw ? YES or NO.
@Vicidominus avra' pure fatto due cazzate mega-galattiche MA resta un grande giocatore e quando si sara' ritrovato (qualcosa lo blocca al momento e solo lui sa cosa) non ce ne sara' per nessuno...
While building OtterForm, I had one of those moments where #AI felt like a superpower and a limitation at the same time. At the beginning it is incredible with features that moved fast, boilerplate disappeared, and it genuinely felt like we had added another developer to the team. But as OtterForm grew into a real product with integrations, edge cases, and long-term architectural decisions, the cracks started to show. AI can write code, but I was still the one holding the system together, acting as the #PM, #architect, #QA, and integration layer all at once.
Every development cycle started to look the same: explain context, get code, restructure it, align it with the rest of the system, and fill the gaps the AI didn’t see. The issue wasn’t intelligence, the models are very strong: it was workflow. So instead of trying to "prompt better" I built something that I actually needed: a way for AI to work like an #AGILE development crew, not a chat assistant.
That system became #CODEMUS. You don’t talk to "AI" directly; you work through a PM role that turns ideas into structured requirements, tasks, and coordinated work across dev-style roles. Less prompt juggling, less chaos, more process. It’s the internal system that helped us keep OtterForm sane as complexity grew, and now we’re turning it into a software running on a #Mac.
It requires #Claude #Code (PRO or MAX). Have a look if it might be of any interest.
👉 https://t.co/V5TyNnCN6a
This project began because I was fed up with constantly bookmarking website after website to remember where I saw that pesky little utility...
Need to convert #JSON? Google it. Need to format some #XML? Google it again. Need to create a #UUID? Once more, you got it... google, click through 3 ad-riddled websites, and discover one that actually does what you want.
So, I built https://t.co/NtuUZd3nBP over the course of several months: I took every tool I could think of and compiled them into one location. Today, there are over 1000 online #tools available on the site including #converters, #formatters, #generators, #encoders, etc.
They are all available for free with no login required and no paywalls; I was just looking for something that worked without any hassles.
Hopefully, some of you who are in my same situation, will also find this tool useful.
If I have missed anything, please let me know since I continue to put additional tools.
I used to build with #AI in the simplest way: one chat, one loop. Ask... Revise... Repeat. It felt fast.
Then I noticed the pattern: the output was often “good”… but shipping still hurt. Edge cases showed up late. UX got patched. Tests were an afterthought. The same work came back twice.
So, using #claudecode, I tried a different setup: I stopped asking for better answers and started running AI like a small team. One PM "talks" to me. The PM pulls in Dev, QA, UX (and whoever else is needed), then ships an integrated slice with clear acceptance criteria. It’s not louder. It’s cleaner. If you want the exact “Crew” prompt I’m using, read the article or comment CREW.
https://t.co/Bsq1SEOMfZ