hanki.dev

build your own tools

Now that LLMs are so good that one can actually build simple apps sooo effortlessly, I've been using it to make personalized tools with my very own specs, for very specific cases. It's pretty cool to be able to just vision the stuff you want, without needing to spend hours coding 'em.

Here are some examples of what I've done. All of these I use actively, and all of these took couple of hours to make:

Pack Buddy screenshot

There are some similar apps available which I was using before I made this, but I had some problems with all of them. For me packing without a list feels like a headache, like I'm so gonna forget something. With lists it's easy and fast; just throw items in your bag. It's summer and you don't need winter stuff? Just skip the 'winter' category from the list! So effortless :P

Echoes screenshot

I'm a quote nerd and been saving random quotes, phrases and sayings for years in a private Telegram group. Works, but my biggest problem is mainly the search, it's not as great as I'd like it to be. So I exported them all, put in an sql db, and made a pretty interface for it all. As well as Telegram bot for adding quotes on the go (since it's hosted locally) and getting random daily quote from the archives!

This one I didn't make public on Github and it made development so much easier. One can just hardcode apikeys and stuff :D Okay I know that's not a good practice, but point being, when you make stuff for yourself and aren't planning to open source it makes the mental barrier of creating lower. Just hack together something usable, with your own very niche requirements, see if it works and if it does, keep using it. Improve it gradually when needed.

Making an open source project, with other potential users in mind (most likely nobody will clone your project) does add more work. That being said, I do want to eventually open source this one ^^

not-plex screenshot

I'm avid Plex/Jellyfin hater. As awesome and cool as they are they aaalways have problems. I've come to the conclusion that for (legally ripped bluray collection) just directly streaming from your server, whether it's locally or not, is much smoother experience.

So my previous workflow was navigating to my server, copying video link, opening Iina, "Open URL", pasting the url and voila! Now I just open this app, click on video link and it opens it in Iina! Many clicks saved, lol. But the best part is the continue watching -feature, it saves last opened links in local storage, and uses regex to check if there are episodes with higher number in that directory.

730am screenshot

Not gonna write much about this as it's so simple: a dashboard for tracking wake up times, and showing cool stats about it. Super simple, hosted on Github pages. Took like 2 hours - and most of that was me trying different UI variations. I do think it looks pre sick now.

I've also made some other little stuff, as well as managed all the self-hosting stuff. Point of this post was to embrace that with coding skills, and the latest ai coding tools like OpenCode, developers can create so much with so little time. I think it's cool. I think it's awesome to create hyperpersonalized stuff just for you, apps which previously would've been too niche to spend time bringing to life.

If you have some random niche diy tools you've made and want to share I'd love to hear about it! Why not create a blog post of it and email me the link ^^

✌🏼 Like my content? Subscribe via RSS feed.