Ask HN: What have you built privately with AI that you'd like to share?

I found that a project of mine grew beyond my ability to reason about many parts of it either individually or holistically and started using AI. AI started becoming a context nightmare shuttling context back and forth from session to session, finding the right context and getting it applied. I eventually came up with Determined. (https://github.com/bartanderson/Determined) I am currently using it for two different projects and I developed it for other people to use besides me but debated showing it to others. AI has been a big help to me in writing it but it came from my realizations through conversations with AI and one pointed one where it told me what I was doing was not done well by AI. So I started doing more deterministic analysis and that grew and grew and I added languages and the concept of fsms used to capture discrete state machine architecture and reason about them as well as coalescing docs and other things about the software. An AI is able to use it pretty easily (I have been using Claude but have been using it to help train opencode to drive a project for it). If you don't like AI you can still use it to analyze your software. It will tell you things that are hard to see unless you wrote it and have a great memory. I doubt I will be taking pull requests for changes, but feel free to pull down a copy for yourself and do with it what you will. That was what I am doing and am happy sharing it now that there was a forum to do so.

Here are some of the things I’ve build recently, all currently private but mostly because I’m scratching my own itch

- Macros: MyFitnessPal replacement, track my macros, meals/presets/custom food items, also pulls from OpenFoodFacts. MCP support so I can tell an agent “Log my normal breakfast” or “log the sweet potato beef meal for dinner with 2 tortillas”

- Foreman: My software factory (https://joshstrange.com/2026/10/12/there-are-many-like-it-bu...). Every one of these is highly custom/personal so it probably won’t be interesting for me to go into a lot of detail. But suffice to say, I am successfully shipping code through my factory for both my side business and my day job.

- Quartermaster: tool to scrape (via a browser extension) my purchase history at various retailers, including Amazon and Kroger (grocery store). Small web UI to view/filter the data, MCP to interact. I can see purchase price over time, quantity over time, and even things like “price per TB over time” for things like hard drives due to a tagging system that lets me slice and dice the data however I want.

- Vendors: a small web app for my mom to track the vendors she recommends (she’s a real estate agent) to clients so that when they ask for a painter, she can go filter down to the painters, select the people she wants to recommend, and copy all of their contact information into a text/email.

- Amdi: my catch-all app/platform for things small enough to not warrant (yet) their own project. This contains my own little chat interface that hooks into all the data sources above and provides rich UI for certain operations. It’s incredibly custom to me and how I think/work. My favorite features are all around my plex library, letting me cut a clip, find a clip (smart subtitle search), find out what else I’ve seen an actor in for the show I’m watching (using my watch history), and a bunch of metadata around my media exposed to the LLM.

I’ve built an SVG animation tool. Something akin to a basic Flash tool. It was originally a rudimentary tool I coded by hand, then used AI to tighten up, fix issues and evolve into something I’m happier with, it’s probably 90% AI now, with features created bit by bit rather than one-shot prompted.

It’s a niche tool for my use; to create animated image assets in a fraction of the file size to a comparable GIF.

There are still issues with preparing work in Illustrator so my next step will be to work ok some form of plug-in that can export prepared artwork for use in the animator.

Im also working on a custom PCB for a software synthesiser using a variety of off the shelf components.

Again niche but hopefully getting the next set of PCBs off to the manufacturer later this week.

I don't feel like I made it, but I asked AI to build a cashflow manager chart thing for me. I can upload CSVs from my bank statements, and it displays my cashflow as a graph over time, and then I can put in upcoming and recurring expenses, so to see how much money I'm going to have to borrow over the next six months until my revenue catches up with all my pending liabilities. It feels like I'm in control if I can see the graph of my money, even if it goes pretty far negative.

A coworker created a whole financials site, and then another coworker did the same thing, but turned it into a game. You add your real financials (age, income, stocks, etc) and then as you walk through the hall it changes your financials and allows you to see the various changes, even has life event changes, retirement age markers, all kinds of ridiculous things.

Ive been working on https://github.com/trypando/pando.

People are building things with AI all the time at work, and with no outlet for it, security violations are running rampant.

Pando gives employees a safe place to put these, and reduces the burden on IT and security teams when dealing with them.

Been working on a self hosted discord alternative for the past 6 months. Comes with most features you would expect of chat software (and a few more). Testing native clients right now https://github.com/Chatter-Org/Chatter

I made an iOS app. It is a metronome and tuner in one screen, useful for me learning violin, so I can play a scale at 60bpm and check pitch. The tuner has to eliminate the tick of the metronome.

I can publish to iOS store if anyone is interested or share source, but you can vibe code your own in an hour for a couple bucks.

Turns out, Opus is pretty good at one-shotting Obsidian plugins - it doesn't even need to be shown a sample - so I "made" a few for myself. (They only make sense in my vault structure, so no use in publishing them, especially when it costs only a few bucks to "make" your own.)

I built a wallpaper manager for macOS.

The sidebar has a tree view of the filesystem, defaulting to where I keep my wallpapers (adjustable in the settings). The wallpapers are shown as thumbnails on the right, size adjustable.

There is an option to show them recursively, so all the wallpapers from subfolders show up in one view, rather than having to navigate the folders. There are options for favorites and a few filters.

There is a preview mode, which is invoked by hitting the space bar, like a custom QuickLook. If previews in full screen, with a mock menu bar and dock, so I can see what it will actually look like before setting it. In this view I can also use the arrows to flip through the wallpapers and use shortcuts to favorite them, or apply it as the wallpaper.

I have some planned updates to it, but thats the core functionality at the moment.

I built an experimental extensible typed array programming language (with tools, libraries, demos, etc. actively under development). Via meta-programming (macros) and Rust shared library extensions it lets you write code that runs on CPU or GPUs (Apple, NVIDIA so far).

> https://blog.softwarewrighter.com/2026/10/16/array-languages...

I originally wrote gpu extensions that use OpenCL/CUDA. I more recently added a modular mojo backend that allows me to use Metal instead of OpenCL on Apple Silicon.

Few weekend micro-projects

1. https://umoji.org/ - find, compose and pick emoji text that you can copy, share or bookmark (I got tired of tracking, popups, ads from various emoji picker websites).

2. https://blug.blog/ - generate static blog and deploy on Cloudflare (I got tired of wrestling with Wordpress and more involved machinery for a very simple task).

3. https://bartist.app/ - real time BART (SF Bay Area) train departure times iOS app.

There are more, some not so micro. Ask again in few weeks/months.

I just picked up a project car, and have decided to use AI as my side-mechanic and helper. I had it build a local site, which has a checklist with subtasks, note and photo abilities and parts tracker. The checklist was built up by ai based on what ive been telling it, etc. I have tuned the project ai to handle part searches, identification, and general help. I just added a trigger for things like adding notes and photos, so that when i add them, a request is fired off to the ai, and it will analyze, id, etc. Ive also given myself a way to directly prompt the ai from the site for help, instructions, diagrams, whatever. Ive got a little thinkcentre pc in the garage where the site is accessible, and its also available on my phone for pictures. Its still early in the car project itself but its already been helpful in identifying the engine, and found 2 issues i overlooked.

Also ai helped me build an offgrid setup for said garage because the garage has no power. Helped me size things out, gave me diagrams, etc. Also has already provided me with a way to setup an automatic generator startup when the battery is low. I havent set that up yet, but its in the pipeline.

I started and then abandoned a really deep circuit simulator, which was going to be able to emulate full actions and code eventually.

aaaand lastly, ive spent the last 2 months in the evenings with ai building a complex rts game in godot, its almost 1.0 ready. Sadly the ai im using has been really struggling to turn out quality 3d models, so thats probably going to be on me, which will delay significantly.

On Mon 16. Nov 2026, version 1.0 of the RTS is publicly downloadable as a build. Every unit and every building in it has a finished 3D model, there are no more placeholders. The assumption is that for you, V 1.0 means that users other than yourself can download the build publicly. Does this describe what you really want to finish, or is it wrong? If wrong, what is the correction?

I built this for my own private use but it’s since become public.

Free to use, no accounts, no subscriptions.

https://trysojourn.app

I was using Claude to bounce ideas for parenting with a focus on biblical principals. Trying to really understand how to discipline my children without provoking them, for example.

I was pleased with the experience but Claude did not make it easy for me to dive deeper into scripture. I wanted a different LLM chatbot experience that was focused on less chatting and more scripture reading.

I TestFlighted it out and friends liked it so I ended up publishing it.

does it exist as a webapp?

Not yet, unfortunately. Nor Android, which people have asked for.

I think Android would be supported first but no specific timeline for it yet.

Its a bit of a SWE bonzai tree, but I built a Spring Application framework I'm using for personal projects https://github.com/Demilich-Digital/henge/tree/main/

Basically, a modulith framework you can fork out to microservices. I don't get to claim to solo dev the next Bing thing, but I have codified and learned a lot of distributed system best practices and explored Spring and Convergent consistency methods.

I built a SuperDisplay alternative for MacOS to use my Samsung tablet as a secondary display — feels just like plugging in a monitor now. I also put together an extremely specialised video streaming host and client to get better quality and stability than Discord's screenshare, alongside an Android TV app to cast this video stream. (It has a bunch of features like media control passthrough and display/window selection on the receiver side; things no existing application implements because of how intrusive it'd be for the sender.)

I built an advanced screenshot tool. I always send a lot of screenshots to Claude, and i find myself moving back and forth between the browser or whatever i am taking screenshots of and Claude

so i made a tool that takes as many screenshots as i'd like and then pastes them at once when i'm ready.

I go down Wikipedia rabbit holes frequently. I thought it would be cool to see diffs of articles over time.

https://memoryhole.app/wiki-timeline/compare?title=Theranos&...

im a mac user and i made 3 ultimate substitutes for app switcher, screenshot tool and stickies - battle tested and mature products from my perspective and i do consider them almost ideal.

would i like to share them? maybe, probably, on thing that keeps me from sharing is the lack of shiny packaging which i almost never have time to put in to.

actually with AI my private projects get to a usable state for the first time :)

started with CodeHydra to work more efficiently with multiple agents: https://codehydra.stho.net/

The second one is a App to share photos during an event: https://snapsync.stho.net/

CRT - Code Review Tool: https://github.com/imron/crt

Built a text based adventure. It works like D&D but with an agent as the GM. Feed it in campaigns and it builds a new one for you to play.