Build your own tools!
I think the first software tool I ever wrote was a scraper for webcomics. It was terrible. I wrote it in C# using Windows Forms in like 2008. But, crucially for me, it scratched the very specific itch that I felt. Since then, I've written all sorts of bits and pieces throughout the years to solve this problem or that.
Make the workflow work for you
I wrote my RSS reader bottomfeeder because I had specific friction when using Miniflux. That's not to say there's anything wrong with Miniflux. There's nothing wrong with it, it's a more polished, better rounded RSS reader than mine.
But I noticed how much of my time I was spending just marking stuff as read as opposed to just reading posts that are interesting.
In writing bottomfeeder, I defined the workflow. Posts roll in and after a time, they disappear. If someone goes on a spree and blows out my feed, I don't need to deal with it, the posts will all disappear into the ether eventually. In using it for the last six months or so I've refined the workflow and it's at a point where it's nice and comfortable for me. In six months time, maybe my workflow will evolve and bottomfeeder will change to accommodate that. That's fine. I don't have shareholders. I don't have users. I can do what I want.
What enshittification?
When you're the primary shareholder, the userbase and the dev team, it's pretty unlikely that the app is going to drop a subscription model on you. Last week I decided I wanted a simple life location tracker app for when I go out trail running so that my wife can check in periodically and make sure I'm not dead in a creek somewhere. Are there apps that do this? Sure. Are there apps that are ostensibly free? Probably. Do I want to deal with the hassle of reading privacy policies and working out who's going to sell me out to ad companies either now or in the future? Not really. So I cobbled something together.
For a sync server, I cannibalised an old web API I wrote for something that never went anywhere to bodge together the bare minimum. For the app side, I reached for Flutter and with the power of a couple late nights, a little AI and a LOT of questions for my very patient front-end team coworkers (Hi Mark!) I pulled together a little app to talk to my sync server.
Now I have a relatively stable tool that I can bust out whenever I want to share my location. Over time the feature set will grow as my requirements change, but crucially, nobody's going to acquire this thing and ask "Hey, what if it had a chatbot?".
The allure of commercialisation
When you tell people about a tool you wrote, inevitably their answer is "Why don't you sell it?"
It's a fair question.
I don't think everything needs to be a million dollar idea though. Sometimes, a $5 idea is fine. Sometimes the satisfaction of logging into your RSS reader on a Saturday morning and going "Shit, I made this" is enough.