I Built My First Serious Project: Captain Hook
TL;DR: I'm building Captain Hook, an AI tool that helps YouTube creators write good video intros, those first crucial seconds that decide whether a viewer stays or bounces (titles and thumbnails matter, but the intro is what usually fails, because creators either give away too much or are simply boring). Working off more than a dozen proven copywriting frameworks, the tool generates an interesting, not-overhyped intro for your video idea, plus helps with titles and the thumbnail concept, but it doesn't come up with the video ideas themselves, because it's a tool for "packaging", not for ideation.
It's my first web service with accounts, a database and payments, built from scratch with Claude's help even though I'm not a programmer. It's also part of a bigger series in which I learn progressively harder things. The goal right now: close out the basic version and hand it to YouTuber friends for testing, to find out whether it solves a real problem at all.
Where the idea for Captain Hook came from
I make YouTube videos and I know the title and thumbnail are the single most important element. Without them nobody clicks, no matter how good the video is inside. That part I don't really struggle with, and besides, there are already plenty of tools on the market that help with it.
But there's another element, just as important and often underrated: the intro, the first 15 to 60 seconds of the video. Once you've lured the viewer in with the title and thumbnail, you have to keep them, otherwise they'll drop off within seconds. YouTube Analytics shows it brutally: the retention graph can fall off a cliff inside the first minute.
The problem is that creators find it hard to write a good intro. We know the entire contents of the video, so we either give away too much up front and the viewer loses curiosity ("ah, so it's going to cover this and this"), or we do it too slowly and too dully. This isn't about clickbait, it's about an honest "hook" (hence the name, Captain Hook) that grabs attention and holds interest for anyone genuinely interested in the topic.
What Captain Hook does, and what it doesn't
The tool helps you write an interesting but not overhyped intro, drawing on more than a dozen proven copywriting frameworks. You give it your video idea, and the AI generates an intro proposal based on it. It also helps you brainstorm title ideas and a thumbnail concept (handy too when testing the three title variants YouTube now allows), so that the title, thumbnail and intro work together rather than repeating each other.
What it doesn't do: it doesn't generate the thumbnails themselves, and it isn't a tool for coming up with video ideas from nothing. It's for someone who already has an idea and wants to "package" it well, not for someone looking for inspiration.
How I built it
I built it with Claude Code and Claude Design (the equivalents of ChatGPT, less known to the general public but more popular among people who code with AI). I'm not a programmer, I only know the basics of HTML and CSS, and this is already a serious service: accounts, logins, an AI API integration (it's Claude that generates the intro from the inputs and the frameworks), and eventually payments too, because using AI costs money, and that will simultaneously verify whether people see value in it.
Before this I'd only built simple static sites (on WordPress or with AI) that displayed content. I'd never made anything that actually processes data. So Captain Hook is on the one hand a tool for myself, and on the other an exercise in programming with AI without knowing how to code. That's why I put so much weight on a very detailed spec at the start, so the AI makes as few mistakes as possible and builds exactly what I planned.
Still ahead of me: the AI integration, refining the spec, testing, and the biggest test of all, whether other YouTubers even see the intro problem the way I do. With titles and thumbnails it's easy to verify in Analytics; with intros there's currently no way to test it directly, and that's where I'm hoping Captain Hook will help.
The wider context, and why I'm doing this
Captain Hook is one of several projects in a series I've provisionally called "Summer of Code". The name will probably change, since summer is ending and I still have a fair few projects to get through. Whatever it ends up being called, the point is the same: I'm learning progressively harder things.
My earlier projects, Focus Space especially, moved around territory that was reasonably familiar to me. Captain Hook is the first one that steps outside that comfort zone. I'd never before built a service with a database, user accounts, data processing and payments. All of that is still ahead of me, so a full summary of what I've learned will only make sense once the project is finished.
What's next
The priority now is to close out the basic version as fast as possible and hand it to a few YouTube creator friends for testing, to see whether they find any value in it and whether they'd want to use it. The first tests will be free, but the service will eventually be paid, because generating content with AI simply costs money, and I think that's understandable.