I Mass-Produced 220 Websites With AI. Here's What Happened.

On February 15, 2026, I made my first git commit. Twenty-eight days later, I had 220 live websites, 647 digital tools, 33 published ebooks, and 255,000 unique visitors. I spent zero dollars on funding. I had no team. I didn't write a single line of code by hand.

This is the honest story of what happened — what worked, what broke, what surprised me, and what I'd tell someone who wants to do the same thing.

The Decision

I'd been watching the vibe coding movement for months. People on X talking about building apps without writing code. Using AI to describe what they wanted and watching it materialize. Most of them were building one thing. I wanted to see what happened if I went all in.

Not one website. Not ten. Hundreds.

The question I was trying to answer: What happens when one person with AI tries to build an entire internet company from scratch in one month?

The Tool That Made It Possible

Claude Code CLI. That's it. That's the entire dev team.

It's a terminal-based AI agent from Anthropic. You describe what you want in plain English. It writes the code, creates the files, runs the commands, commits to git, and deploys. It has full access to your filesystem. It can manage existing codebases. It can launch parallel agents to work on multiple tasks simultaneously.

Before this tool existed, building 220 websites would require 15-20 developers, six months, and $2-5 million. I did it in 28 days from my apartment in Chicago for the cost of a Netflix subscription.

Week 1: The Foundation

Day 1 was spunk.bet — a free provably fair crypto casino powered by Bitcoin runes. Ten games. Dice, coinflip, plinko, mines, crash, wheel, keno, limbo, hi-lo, and a rune tower. All provably fair using HMAC-SHA256 verification. All built in a single day.

That first site taught me the workflow: describe the vision, let Claude build it, review, iterate, deploy. By day three, I was averaging 5-6 sites per day. By the end of week one, I had 35 sites live.

The secret wasn't speed. It was patterns. Once Claude understood my brand (dark theme, orange accent, Chicago-based, SPUNK LLC), every new site started from that foundation. The AI remembered. Each site got easier.

Week 2: The Tool Empire

This is when spunk.codes was born. I wanted to build the largest collection of free browser-based tools on the internet. No signup. No email capture. No paywall for the basics. Just tools that work.

647 tools later, it's the crown jewel of the empire. SEO tools, AI writing assistants, image editors, developer utilities, business calculators, social media generators. 317 completely free. 330 premium for power users.

The tools were built using reusable HTML templates. Once I had the pattern for a tool page (dark theme, input area, output area, save/copy buttons), Claude could generate variations faster than I could come up with ideas.

Week 3: The Meme Generator

Clown.best became the breakout product. A free meme generator with 777 templates, a GIF factory, and something no competitor has: Clown Score AI — a tool that rates any X/Twitter profile on 13 categories and generates shareable score cards.

Why memes? Because memes are marketing. Every meme shared with the clown.best watermark is a free ad. Every Clown Score card shared on X drives traffic back. The product markets itself.

The pricing: free tier with watermark, Pro at $9.99/month (cheaper than Imgflip at $10, Canva at $13, Kapwing at $16), and an invite-only SPUNK 13 tier for the inner circle.

Week 4: Scale and Revenue

By the end of the month, the numbers were real:

All verified by Cloudflare analytics. No fake numbers. No inflated stats. That's a core principle: facts only.

Revenue started trickling in. Google AdSense across all sites. Affiliate partnerships with Amazon, Coinbase, and Ledger. The first Bitcoin payment came through the Clown Wall — a feature where anyone can post a message on a public billboard for $0.99-$99. First buyer: KAYY. 55,330 sats. Forever legendary.

What Broke

I'm not going to pretend this was smooth. Here's what went wrong:

• GitHub flagged my account. 227 repos pushed in rapid succession triggered some kind of anti-abuse system. All 220 GitHub Pages sites went 404. Took days to migrate to Cloudflare Workers as a backup. Lesson: don't put all your hosting in one basket.
• Duplicate content hurt SEO. Some sites had 4-6 blog posts targeting the same keyword. Google noticed. I had to consolidate duplicates and redirect. AI can write a lot of content fast — but it needs human editorial judgment.
• The checkout was broken. I shipped a checkout page that gave instant access when someone clicked "I paid" without verifying payment. A user caught it. Embarrassing. Fixed it to require manual verification. Lesson: test every purchase flow as a buyer before shipping.
• Mobile was rough. When you're shipping 8 sites a day, mobile optimization gets rushed. Several sites had broken layouts on iPhone. Had to go back and do a mobile-first audit across everything.

What I'd Do Differently

If I started over tomorrow:

1. Focus on 20 sites deeply, not 220 broadly. The 80/20 rule is real. clown.best, spunk.codes, and spunk.bet generate 90% of the value. The other 217 sites are mostly SEO outposts.
2. Add Stripe from day one. PayPal and crypto are not enough. 70% of buyers want to pay with a credit card. I'm still missing those conversions.
3. Build the email list from day one. 255K visitors and I captured almost none of their emails. That's the biggest missed opportunity.
4. Ship the ebooks on Gumroad immediately. 33 ebooks sitting on my own site when Gumroad has a built-in audience of millions.

What I'd Tell You

If you're reading this and thinking about vibe coding your own empire, here's what I know for certain:

The tools are free. The hosting is free. The AI is $20/month. The only thing stopping you is the decision to start.

You don't need to build 220 sites. Build one. Make it good. Make it useful. Ship it today. Then build the next one tomorrow.

The vibe coding revolution isn't about quantity. It's about removing the barrier between having an idea and making it real. That barrier used to be years of learning to code. Now it's a conversation with an AI.

Start the conversation.