Builder Journal: The Site Engine I Never Stopped Rebuilding

I didn’t invent the idea of a “site engine” recently.

I’ve been circling the same concept for years, sometimes building it myself, sometimes outsourcing it to a CMS, sometimes just avoiding new site ideas because I didn’t want to pay the setup tax again.

This time, I finally shipped the version I actually wanted, and the way I shipped it surprised even me.

The original itch: “why are we building sites by page?”

A long time ago, I learned that many people priced websites by page.

That never sat right with me, because most sites are the same shapes repeated:

• shared layout

• shared navigation

• shared styling

• different content

If you understand databases, you naturally want to separate structure from content.

So I did, first in PHP, later more cleanly with CodeIgniter. The idea worked. I built sites that way.

Then real life happened, client needs happened, and I did what a lot of practical people do.

I moved to WordPress.

The long detour: WordPress solved “publish”, but not “start”

WordPress (and builders like Elementor or Avada) can absolutely ship real sites.

I still have sites running there today, including client work and my own projects.

But it never solved the part that drained me the most:

Starting a new site still felt like paying a tax.

Not hard, just repetitive:

• baseline structure

• blog wiring

• media setup

• SEO basics

• theme tweaks

• publishing rituals

That’s how “someday” sites become placeholders.

The nudge that mattered: watching creators get stuck

In a marketing class for music producers, I watched creative people hit the website module and stall out.

Tools often force a bad trade:

• easy, but everyone ends up with the same vibe

• flexible, but now you’re doing technical labor instead of creative work

That was the signal I couldn’t ignore.

People don’t need more generic templates, they need a system that gives them momentum while still leaving room for personality.

This time I used ShipKit.ai, and I didn’t write the code

Here’s the key difference with Headstring Web.

I built it from a ShipKit.ai template, the kind of starter that’s designed around AI workflows and background tasks, especially for things like long-running jobs and processing. 

Then I leaned into a workflow that felt less like “coding” and more like “directing”:

1. ShipKit gave me a strong starting structure

2. I used AI-driven tasks to add features step-by-step

3. I mostly did verification:

   • does it compile?

   • does it behave correctly?

   • does the data model hold up?

   • does the UI match the intent?

   • do edge cases break it?

For this entire project, I did not hand write a single line of code.

I made decisions, reviewed what got generated, tested it, and kept moving.

That experience made a quote click for me that I had seen online, and it stopped sounding crazy.

Eden Marco wrote on LinkedIn: “I haven’t written a line of code in over a year,” in the context of using coding agents differently, more like delegated systems than copilots. 

I’m not claiming I’m living that life forever, but I now understand how someone gets there.

What Headstring Web is, in practical terms

The backbone is still the same idea I had years ago, just modern and visual:

Sites → Pages → Sections → Blocks

From there, the tool became a real daily driver:

• Visual building with 18 blocks and lots of templates

• Themes via CSS variables, plus AI theme generation to skip blank-canvas paralysis

• Blog system that feels native, not bolted-on

• Publishing rules that make draft vs published predictable

• Custom domains through Vercel, plus instant hosting at /sites/[slug]

And yes, this site lives inside the engine, which keeps me honest. If something feels clunky, I feel it immediately.

Why I think this could help others, even if I built it for me

First, it’s for me. It stops me from rebuilding the same foundation over and over.

But I also think it fits a bigger group, especially creators:

• people who want structure that prevents chaos

• people who still want their site to feel unique

• people who get stuck at “setup” and never reach “publish”

The best compliment I can imagine is someone saying: “I finally shipped my site, and it still feels like mine.”