New post: My week with Claude Code – hard work and gotchas https://www.sarah-savage.com/my-week-with-claude-code-hard-work-and-gotchas/
My week with Claude Code – hard work and gotchas - Sarah Savage

I'm a firm believer in the idea that you have to use a tool to fully critique it. With that in mind I committed to using Claude Code for an entire week, putting it through its paces to see what it could - and couldn't - do. First, the numbers: I used Claude Code for

Sarah Savage - A fusion of development and business
@sarah Brave setup to test AI in production.
@carstingaxion I reviewed and applied every change manually. I only gave it read-only.
@sarah Yes, I did read your article before I commented. And I meant that without irony, I’m afraid of doing such a test, because I might miss one important thing in such a wave of lines of code.
@carstingaxion you get a lot more mileage if you demand the AI justify its behavior and decisions before you apppy them.
@sarah Really interesting to read. Thanks, Sarah.
@ross you’re welcome.
@sarah thanks for this really interesting article. Where does Claude show you the equivalent human hours? Did you feel it really was 4 months saved?
@DaveLiddament I asked Claude to call estimate conservatively based on the notes that it made in my note taker. And yes, I do believe I saved months worth of manual work.
@sarah thank you, this is a very interesting and thorough take!