Claude Code in action

I'm kinda cheap. I don't like spending money on things, in general. So it was hard for me to pay the $20 for the Pro tier, but really, its so worth it. I'm a big time clicky-clicky, IDE guy, but the ability to have it access outside of the IDE is pretty unbeatable. To be honest, that concept for me was introduced by the good people at Warp.

Another really awesome feature is that it has a little task management engine built in. For larger complex tasks, you have to break things down for the agent, use TDD, or you can use external tools (like Taskmaster AI). I've gotten a lot of tips from the DevOps Toolkit channel, and lots of helpful information about techniques and tools, like MCP servers, that can help you improve your model's abilities. I have a post about building a task management system for my model and then found that Claude Code was already doing this on its own (now I realize it can be risky to chain yourself to a particular agent, but for now, it's very nice).

For the most part though, it's the quality of the code it writes. It is just better, and it is better at getting it right the first time. That is a bit of a loaded statement. I mean that it plans it out well and then proceeds. Actually, my biggest complaint with most of the agents is that they try to one-shot all the things.

No...you don't.

Me to lesser agent: "Stop. Don't. Make a plan. Then proceed. And at every step, write a test and tell me how it went (pats head)." You should always make a PRD, and with Claude Code, it kind of already starts with that mindset and it makes ALL the difference.

So far I'm really enjoying it and it is basically a force multiplier for me and has allowed me to actually start a lot of projects that I've been wanting to. In fact, I'd say so far, it has made me start to look at other people's problems in a new light, or actually want to seek out what people's issues are that could be solved with technology. Hopefully, I'll get to churn out something useful 😆.