Well, holy crap. 😇👻💩

Day one of really diving into getting Claude set up as one of the cores of my AI development workflow, and somehow my workspace, my workflow, and my project feedback loop all got permanently better in a single afternoon.

I’m 43. I’ve been doing this for a long time. Which means some of my habits are… let’s say durable.  I've also been unknowingly suffering from a classic case of developer workflow friction caused by constant context switching. Today, a few of those issues finally died the deaths they deserved.

Three things changed... fast:

1. Physical workspace friction vanished

2. An outdated workflow got a real overhaul
3. My project management feedback loop tightened dramatically

The problem I didn’t realize I had was friction I’d normalized. 

When I’m developing, my actual dev environment lives on a MacBook. My comfortable setup — chair, keyboard, monitors — lives on my PC. That meant context switching. Sliding laptops around. Typing on a keyboard I don’t love. Mentally juggling where “work” was happening at any given moment.

I dealt with it for WAY too long.

What made this impossible to ignore was Claude. Once I started seriously using it, I wanted it embedded directly in my real dev environment — running tests, executing shell scripts, touching the same code and services I was working on. But I also wanted to stay planted in my preferred seat, on my preferred keyboard, with my normal monitors.

That tension finally forced the issue.

Enter VS Code and the Remote – SSH extension.

After a bit of setup (and giving the MacBook a static IP so I’d stop chasing configs), everything clicked into place. I’m now working from my PC, inside VS Code, directly against the MacBook’s environment. Terminals are running exactly what they would locally on the Mac. Services are running where they belong. It's a remote development workflow, even though the target machine is on my desk.

And the truly wild part: port forwarding just works.

localhost:4000 on my Windows machine maps straight through to localhost:4000 on the MacBook. No hacks. No ceremony. No “wait, which machine is this running on?” moments. It’s just… there.

Claude lives right inside that environment now. It can run tests. Execute scripts. Inspect results. All without me bouncing between machines or breaking flow. I stay put. The environment stays consistent. The friction is gone.

That alone would have been a huge win — something I honestly wish I’d sorted out a decade ago. I probably never would have if Claude (and ChatGPT, frankly) hadn’t casually nudged me in this direction.

But then came the icing.

While wiring Claude up, I connected it to my Atlassian account. Jira. Confluence. Suddenly, it can interact directly with my project management flow and documentation. It’s already generating copy, updating tickets, linking docs, and keeping things in sync automatically.

The amount of time I spend filling out ticket details and writing documentation is about to drop dramatically.

This wasn’t just a tooling tweak. It was removing a foundational constraint I’d quietly accepted for years — and once that snapped into place, everything else stacked on top of it effortlessly.

HOLY CRAP.