Why I Built jdavies.io

I’ve been writing Security Signal for a while now, and posting on LinkedIn three days a week. Both do their jobs well. Neither one is mine.

Substack owns the newsletter. LinkedIn owns the feed. And every time I shipped something on the builder side, a print farm app, a tool that catalogs 3D models on its own, a Pokémon card checklist I absolutely did not need to build, it had nowhere to go. It’s not a threat breakdown, so it doesn’t belong in Security Signal. And a LinkedIn post can’t hold a real build log.

So I built a home base.

What goes where

Security Signal stays the place for threat intel and AI security. Same cadence, same format. If you haven’t subscribed, go do that.

This blog is the workshop. Build logs, experiments, the things I tried with AI that worked and the ones that fell over. Less polished, more honest, closer to the notes I’d keep anyway.

How it’s built

In the spirit of building in public, here’s the stack. It’s an Astro site, which means plain static HTML by the time it ships, running on Cloudflare. I push to GitHub and it deploys. Hosting cost is zero. The latest Security Signal posts get pulled in at build time, so the homepage stays current without me touching it.

I built it with Claude Code as a pair programmer, from an empty folder to a live site in one session. That’s the part that still gets me. Not that AI can write code, you knew that already, but that the gap between “I should build that” and “it’s live” keeps shrinking.

AND THAT’S THE PART THAT MATTERS.

If one person can ship this fast, the interesting question isn’t whether you can build something. It’s what you should build, and how you secure it once it exists. That’s the overlap I want to write about here.

Stick around. There’s a lot more coming.