Six Months on Claude Code: A Mid-Year Reset

Six months ago, I made a quiet decision. I would run my entire operation on a small set of AI tools and see what broke.

Today I am sitting at the mid-year mark, and very little broke. What did happen surprised me.

I want to share what I learned, less because I am proud of the system, and more because the people I work with keep asking me the same question. What do you actually use, and is it worth it?

Here is the honest version.

What Earned Its Seat

Four tools made it to June with me, and I am not adding any new ones until at least Q3.

Claude Code is my development environment. It is the difference between "I wish someone would build this for me" and "I built it this afternoon." Internal tools, scrapers, dashboards, scripts. Things I would have paid an engineer five figures for, I now build between meetings.

Claude is my writing partner and my strategy thinking partner. Every article, every deck, every cold email gets workshopped here first. I never let it send raw output. I edit every line.

Gemini is my second opinion. Different model, different biases. When I want triangulation on a hard call, this is the room I walk into.

GitHub is the boring hero. Every project, every script, every iteration. Future me thanks past me every single time.

That is the stack. Four tools.

What I Cut, and Why

The harder lesson is what I stopped using. I cancelled three subscriptions in the last 90 days, all of which I had convinced myself were essential.

One was a workflow automation platform I touched twice in three months. One was a fancier analytics tool I never opened. One was a "second brain" app that mostly became a graveyard for screenshots.

The pattern was the same in all three cases. I had bought the tool for the version of myself I wanted to be, not the version of myself I actually was.

The Discipline That Mattered More

Here is the part nobody talks about. The stack is not the thing.

What moved the needle in H1 was not the tools. It was the daily discipline of opening one of them, every morning, before email and Slack could pull me into other people's priorities.

Tools compound when you use them every day. They sit unused when you treat them as something you will get to later.

What This Means for You

If you take one thing into the rest of the year, pick the smallest possible stack that you will actually open every morning. The rest is noise.

I built a free H1 AI Stack Audit you can run on your own stack in about 30 minutes. Same filter I use on myself. It is linked below.