
After a year with Claude Code as my daily driver, the Codex desktop app's interface for managing many agent threads and automations has me reaching for it more and more.
I'm increasingly reaching for the Codex desktop app as my "daily driver" AI agent tool - for software development and day to day automations and "loops".
I've been a huge Claude Code fan and have used it as my daily driver since May of 25. I am still reaching for it in "new to me" code projects where the agent based lifecycle is less defined and when planning out bigger product efforts, etc.
But the interface to the Codex desktop app is just way better for managing all of my many agent threads as well as automations. The automation capability which is increasingly part of my day to day is just more flexible and better right now. The Claude desktop app's interface is a mess. I keep it mostly as a purely Cowork interface now and keep Claude Code in the terminal.
GPT 5.5 is still more "robotic" feeling than Opus, et. al but a lot of what I'm doing is just rote getting things done and it excels there as well as with tough technical challenges. It does a decent job of being an editor.
I'm happy to keep paying for my max plan to mix in - I still have each one review the other.
But until Fable comes back, I'm back in the OpenAI badged driver seat.
Get More Like This
Follow along as I build and share what I learn
Found this helpful? Share it with your network!