By Eric J Gruber
Published on: 2026-02-20
Over the course of a two-week period in December 2025, I had some time off from the day job so I decided to work on a little personal coding project.
The website I created, Delightfully Aggressive, is something that has been on my project list for some time. I decided it was time to make it real.
I have used a few different AI product over the last few years, but ultimately I keep coming back to Claude by Anthropic. Claude seems to be the best at code generation and thinking like a developer, but there’s also the issue of trust: I trust Claude to do right by those of us who write code.
Perhaps that trust is misplaced, but that’s where I am today.
To begin, I started my project using Claude with the most important step: planning. I began by creating a Project in Claude, and then gave it this as a descriptor of what I wanted to build:
I am a fan of music, and I would like to create a personal website that tracks available concerts in my area for the genres I like. This site will use Vue.js and Astro, mainly.
From there, I told Claude I wanted to get data by calling any free APIs to start that would give me results of concerts in my area (up to a 60 miles radius), and limited to the genre of metal, but also including sub-genres to throw the widest net. Claude provided me a list of APIs from music services (Songkick, Bandsintown, Ticketmaster), in addition to creating a web scraper for websites I know of area venues that have shows I like, but don’t provide an API. Claude also gave a third option I already considered but is less available these days: pulling from an RSS feed.
I decided to go with the Ticketmaster API to start, but also have a way to add items manually. Start small, build from there.
After that, I asked Claude to create for me a plan to get to MVP, which it provided in chunks to work on throughout the duration of vacation. This is a far cry from doing professional work with a large team on a big, enterprisey project, but it was a great way to get started for a solo project.
After setting up my local development environment on my laptop, I was off to the races. For a couple hours a few days a week, I worked on my little project.
In many ways, I didn’t consider the experience to be much different than using a search engine to get answers to the questions I had, with the expection that it helped generate code based on my instructions and feedback. I find most search engines to be terrible these days, and paying $20 a month to get a personal assistant to chat with and get decent “search” results is well worth it, in my opinion.
And so, we coded away, just Claude and I. I gave input, it gave output. I ran into errors, shared the errors with Claude, and it helped me work through them. It even generated a basic design for me to use, and I generated a logo that was good enough to launch with.
After looking my repo up with Netlify, adding a security certificate, and pointing my domain to the right location, I had something to show the world.
The site went live on Jan. 1, 2026.
The time I spent working on this was nothing but pure fun for me. I enjoyed interacting with Claude, a virtual “rubber duckie” if you will, and really benefitted from the ability to have it explain something to me that I didn’t understand. I think it’s important we don’t lose human enrichment as we use these tools, and having it spit out code and shipping it without deeper understanding doesn’t help when future bugs arise, or for testing and long-term development.
There are some bugs I have found post launch, and other things I want to continually improve upon as the site grows. But overall I am pleased with the site, and am pleased with what Claude has to offer in ways of a little coding buddy. I look forward to working more on my website and seeing where it goes.