Commit

2e7100f

Tempted to change my dev setup every week at this point. New harnesses, new models, agents running on a vps, agents running in the cloud, yet another app to manage it all. How are you handling this? currently using tmux and Herdr to run my clankers on a Mac mini, but still…
If it works, don't break it... I keep getting the urge to try new things, but it's just a distraction from shipping for me! I'm currently on Solo on my MacBook Air and Herd. Claude and Codex with some robust "how we do things" scripts and rules in my agents.md. I have a Mac Mini I've been doing lots of my larger regression tests on and that's great to save CPU/Memory for all the agents running. My advice: you'll never be happy, only change one thing at a time. I just had to move to Zen from Arc and it's the only thing I'll change for a week or so. Then I'll probably make a small change, then see how things go. Small incremental improvements...

commit 2e7100f

Yes, good to reflect before doing any change; and if, then trying it out deliberately. And embracing the fact that i'll probably always have a few different ways, rather than the one ideal one