Claude Code Can Now Build a Real Website on Its Own Over Several Days

Opus 5.5 in Claude Code is exciting. Not exciting in the way a new gadget is exciting for a week and then forgotten. Exciting in the way that makes me sit down and redo the math on a business.

The biggest change is simple: it can run multi-day tasks now.

What I actually did with it

I put it to work upgrading the birdsittingtoronto.ca website. I gave it one prompt, and that prompt ran across three to four days. It did an amazing job.

The new site looks way better. It's faster. It has better SEO. It's more informative, and it's more functional. On every point that matters for a business website, it came out ahead of what we had before.

It's also off of WordPress now. That alone is a big deal to me.

Why multi-day matters

A tool that can help you for a few minutes at a time is useful. A tool that can keep working on the same job for days, and finish it well, is a different thing entirely. That's the line Claude Code has crossed.

Building a proper website isn't a five-minute job. It takes sustained work on the same project. Now that the tool can carry a task across several days, the whole job can be handed over, not just small pieces of it.

The numbers

Here's where it gets interesting from a business point of view. We can autonomously make $5,000 to $10,000 websites now with Claude Code.

As a single person, I would be able to make ten websites like this in parallel.

At $5,000 a website, that's roughly $50,000 a week, and I'm being conservative. I'm using the bottom of the range on purpose.

I run a web agency, so I'm not saying this as an outsider guessing at what websites are worth. A site at the level of what was built for birdsittingtoronto.ca is the kind of work clients pay real money for. The difference now is how much of it one person can deliver at the same time.

There are so many opportunities out there

What I keep coming back to is that this isn't only about websites. Websites are just the example I've now seen with my own eyes, on my own business, with my own prompt.

When one person can run ten serious projects in parallel, the ceiling on what a small operator can do moves up. A lot of businesses were built on the assumption that output scales with headcount. That assumption is getting weaker.

There are so many opportunities out there. The people who sit down and actually use these tools on real work, not just read about them, are going to see them first.