using codex after hitting claude code limits feels like going back to a shitty ex

I just ran down my $200 Claude plan in a week and this is what I learned. Get familiar with vim even if you're not intending to be a software engineer. You can just use Claude code in a terminal only which saves memory. And I also recommend using ghosty so you can have multiple tabs open. This is what the creator of Claude code uses as like the way he does **** right now too. I taught Claude code how to offload making prompts to ChatGPT because as good as Claude code is to code, like anything ChatGPT is, is able to write prompts better. But all you got to do is just make sure you give your clawed markdown file as contacts. But all you got to do is just make sure you give your clawed markdown file as context. So then you can have multiple clawed codes working together while ChatGPT is drafting the next set of prompts for them to work on. And three skills and stop hooks. These two things are all you need to know. Skills are just ways to repackage **** that you just make claw do over and over again. And knowing exactly how to control your stop hook, which is when clawed code stops its agentic loop. It just saves you so much time.