alt Hacker NewsAsk HN: What are you working on? (January 2026)
What are you working on? Any new ideas that you're thinking about?
Comments
I am building an open-source, cross-platform desktop app for building and providing code-related context to chat based large language models. [1].
I found that sometimes I would rather interact with a chat interface to debug an issue or brainstorm architecture solutions in my repos. Agents are great for giving the model access to everything and letting it figure it out.
By manually prompting, it forces me to keep my metal model of the codebase up to date, and it allows me to provide just the context I want to the LLM.
I have always assumed this is about individual projects, but let me try talking about our team effort. We just launched exe.dev, and have lots of projects. In particular, we are spending a lot of time on file systems. The easy one is copy-on-write VM cloning, which existing software provides for us. More interestingly, I believe the standard cloud approach of putting everything on a NAS by default is wrong. Doing better here is going to require doing unusual things.
Finally, we decided to open source the exe.dev agent, Shelley. https://github.com/boldsoftware/shelley
During the holidays, I’ve experimented with some ideas. I wondered if it was possible to make money in 2025 using simple, nano banana wrappers, thr answer is yes!
for example, RecolorLife.com and Headshoti.com generate around $800 USD.
Now I will expand for real estate.
Warenkorb+ (German for „shopping cart“): https://github.com/Lars147/warenkorb_plus
A browser extension (Chrome & Firefox) for simplifying my online grocery purchase workflow from Cookidoo to Knuspr.
I was tired of my weekly workflow of copying, pasting & sorting the grocery page for each item.
Also launched my first Hugo blog. Really nice experience so far. Wrote more detailed about the extension as my first blog entry: https://lars147.github.io/blog/
Recently have been focusing on a personal assistant type of thing.
I've been building it with the agent sdk and any time I want an additional skill, I create it
Examples: parse this pdf containing my credit card bill and add all transactions
Given it has a db, I've been using it to save notes, ideas etc.
Been fun
I've been building Nottawa for ~4y now and it's finally out in the open!
Nottawa's a free macOS app for making live audioreactive visuals. I'm trying to position it as a 100% free, batteries-included alternative to Resolume and TouchDesigner.
Not a tonne of users yet, but I'm hoping to get some traction in 2026. Would love love love to hear some feedback!
1. Zig-clr. "Borrow checking and other safety analyzer for zig"
https://github.com/ityonemo/clr
2. Molecular biology editing software. Will plug in to agentic ai workflows
Been working on canine.sh for about 2 years now.
It’s an open source project that basically turns your kubernetes into a developer friendly PaaS.
Just crossed 2k apps on the cloud version, no idea how many people run it locally, and thanks to a generous sponsorship from the Portainer folks, I’m able to work on it close to full time.
A system-agnostic language for magic spells with a compiler capable of producing the magic wand movements, incantation, hand signs or magic circle required to perform the spell
Some time-constraints make it hard for me to commit to some open source work right now, but if I can I still improve my tool called "Software Manager". Goal for 2026 will be to have registered 4000 programs in total (I am at about 3750 right now) as well as make it dead-easy to interface with it through oldschool SQL (right now I just have .yml files and other flat files, though I can autogenerate a .sql database already as-is, but right now I don't query this dataset and it is not really optimised either). My long term goal is to make compilation on any platform dead simple. On Linux it works very well; on windows not quite as well yet.
An iOS app made for saving locations. It’s called Pyxis: https://apps.apple.com/us/app/travel-journal-pyxis/id6639617...
I’m not a developer by trade but I’ve been learning iOS dev for about 6 years now. It’s become my project that I just keep working at since I personally use it a lot.
The app lets you save your favorite locations, add notes to them, add photos, check weather, tag them for better organization, and archive those tags for future trips. You can also mark off locations that you’ve been to already: think breweries or a coffee shop when visiting a new city.
For the next update, I’m working on a task list functionality for each location. The idea came as a shopping list based on which stores I go to but it can work for any other context as well. This way I can get rid of my shopping list from my task apps.
In terms of weather, I’m also adding historical averages to the forecast to have some sort of context to the weather.
Also leaning more into marketing these days (hence this post) and designing a new icon with some custom art work to give the product some sort of personality. I started learning affinity design to just do it myself so I learn some design software along the way.
Anyways, if you download it, I’d love to hear some feedback. :)
Building CodeIQ - an AI tool that automates transaction coding for accountants and bookkeepers.
The interesting technical bit: it analyses your historic general ledger to reverse-engineer how you specifically categorise transactions. So instead of generic rules, it learns your firm's actual patterns - "oh, they always code Costa Coffee to Staff Welfare, not Refreshments" - that kind of thing.
Posts directly to Xero, QuickBooks, Sage, and Pandle. The VAT handling turned out to be surprisingly gnarly (UK tax rules are... something).
Been working on it about 6 months now. Still figuring out the right balance between automation confidence and "just flag this for human review".
I built an experimental system to test whether an autonomous AI can propose actions freely but be structurally prevented from executing side-effectful actions without explicit authorization. The system consistently blocks unsafe filesystem, shell, and network operations and produces a trace and diff proving nothing changed, even under adversarial pressure. The goal was to see if refusal and non-action can be enforced and verified at the execution layer rather than relying on prompts or logging.
Over the holidays I built MakersHub.dev – a community platform for developers and creators building with AI tools. Whether you're a complete beginner exploring AI coding for the first time or an experienced professional showcasing advanced work, the goal is to have something useful: learning guides, project showcases, discussions, a tools directory, and news. Still early but actively building it out. https://makershub.dev
Just released a Code Quality leaderboard based on the static analysis research I've been doing. https://codehealth.sibylline.dev/
Now working on comprehensive benchmarks for another tool I built, https://github.com/sibyllinesoft/scribe. Results thus far showing it reducing agent token usage by ~80% in real world tasks, but I need to repeat to get variance. Hopefully I can get a writeup out soon.
Developing a graph-based programming language called Graphoid. The original concept was to target AI-related programming. Probably not ready for that yet!
Repo: https://github.com/xvandervort/graphoid
Claude Code is doing the majority of the coding, with close supervision from me. I write notes while I'm working on it. Notes are here: https://www.patreon.com/cw/aiconfessions
2 projects around virtual tech events:
Open Tech Calendar, listing virtual tech events that include community participation: https://opentechcalendar.co.uk/
A listings site for virtually attending FOSDEM. The live streaming is great but the official site only lists sessions in the local Brussels time zone. You can choose your time zone here: https://virtuallyattend.teacaketech.scot/fosdem/2026/
I 3D printed some new apertures for an infrared spectrometer at work to reduce some issues that cause artifacts in the data to enable higher accuracy measurements, particularly of high index and reflective samples. Seems to be working well now.
I've been working on binding raylib to s7 scheme and implementing a kanren on top of it basically just for the sake of understanding. I let AI write most of the binding code, though, because that is conceptually simple but very boring.
CodjiFlo[1,2] - a code review tool inspired by Microsoft's CodeFlow, used by ~40,000 developers. It is especially tailored to power users of pull requests to improve contextual understanding and ease of code review and collaboration.
OpenRailwayMap is a project focused on displaying everything railway related in the world, powered by OpenStreetMap data.
I am working on Entangle https://entangle.cloud a way to add AI powered chatbot to a website.
I looking mainly for European market that is willing to self host your content or the model on your own servers, we will provide the technical support in doing so that and easily plugin the AI powered search or chatbot to the website.
But it is also possible to use servers provided by us to host the content and the LLM models.
We're doing so much planning and reviewing with coding agents like Claude Code and OpenCode.
I spent a day over break building a better UX for reviewing coding agent plans.
Plannotator - Annotate and review coding agent plans visually, share with your team, send feedback to the agents with one click.
Demo video: https://www.youtube.com/watch?v=a_AT7cEN_9I
I have been working on Print Relay[1], a cloud-based print relay system that enables remote printing to any printer connected to your local network.
PrintRelay consists of a cloud server and lightweight clients that connect printers to the cloud via WebSocket.
We use the excellent SaaS PrintNode at work, but about twice a year we have connectivity/routing issues between AWS ap-southeast-2 and their servers OS. PrintRelay is my attempt to not need PrintNode. Because of this PrintRelay is PrintNode API compatible.
I built the running app I always wanted: https://runcoach.fly.dev
You get tailored running schedules and also some body weight strength workouts and healthy meals all in one!
I am a passionate player of this traditional German card game called "Doppelkopf". It's a fantastic way to spend time away from screens. The game is quite complex which makes it super fun but also hard to pick up.
I'm building a small web app with an interactive tutorial and a browser-based singleplayer game that helps people learn and practice Doppelkopf. I've just released an English version:
I’m working on a context aware clipboard for Mac.
I found myself switching a lot between apps to get the same info, lots of copy/pasting.
Example, URLs in bookmark (which I forget about), project descriptions , images, folders.
So I built a Mac app that is similar to Raycast, but just for notes. If I want to save a webpage, I click control+option+C and then a window pops up to describe it.
If I press control+option+V, I get a spotlight like window where it does full-text search of all my notes and descriptions and filter so I can either:
- Open
- Insert the data into the current app (chrome, slack, ChatGPT).
I’ve been using it for a few weeks now, and not sure if others will find it useful.
Just got started again on my "Poor Man's Bitemporal Data System" (discussed here: https://news.ycombinator.com/item?id=45118585 ).
Holidays nuked all the hot-cached context in my head. I spent a few days just spinning wheels until it repopulated. But the basic idea works now!
Much testing and benchmarking work remains to make sure it's not going to lose data, and that it won't denial-of-service itself (because object-map -> facts fan-out is big).
Also a second giant blog post is due (following the one discussed above). Lots of notes have accumulated.
It will be fun even if the concept ultimately crashes and burns to the ground :)
In which case, there's always datomic and xtdb :D
(def repl-facts
[{:alt.site.evalapply/meta
{:description "Root namespace for a named site."}}
{:alt.site.evalapply/features
{:paid #{:alt.site-feature/feat-1
:alt.site-feature/feat-2
:alt.site-feature/feat-3}
:trial #{:alt.site-feature/feat-4}
:complimentary #{}
:available #{}}}
{:alt.site.evalapply/users
{:authorised #{:alt.user.evalapply/user-1
:alt.user.evalapply/user-2
:alt.user.evalapply/user-3}
:unauthorized #{:alt.root.*}}
:alt.user.evalapply/user-1 {:name {:first "Wiley"
:last "Coyote"}
:roles #{:alt.role.evalapply/owner}}
:alt.role.evalapply/owner {:rw #{:alt.user.evalapply/*}
:ro #{}}}])
(assert-facts! (#'user/system-state)
::app
:alt.root/user-1
(uuid/v7)
repl-facts)
(redact-facts! (#'user/system-state)
::app
:alt.root/user-1
(uuid/v7)
(subvec repl-facts 2))
(read-now! (#'user/system-state)
::app)
I'm getting ready to release a re-creation of an old mobile game that got delisted after the devs sold it to a company that subsequently ruined it.
The stack is Babylon + React + Capacitor, which was easy to step into as a full-stack dev with zero game building experience. Currently seeing what I can do to fix some performance issues, though it still works decently for a graphics heavy incremental/idle game.
Beta is still open for Test Flight. Can sign up via https://blossom-beta.blindsignals.io
I'm working on interactive explorations of algorithms and machine learning. Small, visual, hands-on demos that help build intuition by letting people tweak code/parameters and see how things behave in real time.
building a GPS tracker for my cats who like to wander outside.
some features:
- no monthly subscriptions
- location via GPS/GNSS
- a screen that hangs on my fridge (akin to marauders map, to see where the cats are at all times)
- the location data stays local always.
The tech will be extended to more products - a watch for adults, kids tracker etc. Will release here once I have all the tests completed!
I'm adding drug-development features, and polish to my open-source molecule viewer and editor, Molchanica: https://www.athanorlab.com/molchanica.
Source code, in rust: https://github.com/David-OConnor/molchanica I've split out its building blocks into their own libraries on crates.io, for anyone building other bio or chem software. I don't think anyone uses them at this time.
For work related items I’ve been building out agent tooling for building some models and PoC projects related to energy industry applications. Been doing the consulting thing for a bit now and gaining more broad knowledge on some of these data center builds. Hoping to spin this into a product soon.
Started playing with gas town which is really cool. I had a naive version built that was just not good enough. This feels like a step in the right direction.
Haven’t had much time to work on any of my physical hands on hobbies lately but maybe when the weather gets better I’ll head back out to the shop again.
Still working away on a free, completely offline transcription app that's available on all platforms with CUDA support as well.
https://blazingbanana.com/work/whistle
Currently tidying up some internal code (also removing the larger model on mobile platforms) and implementing proper diarization (who said what) so that it can be used for more than just personal dictation.
My iOS developer account is _finally_ approved so it will be available through the proper app store soon.
A domain-agnostic, open source scientific database.
“Protein Data Bank-in-a-box”
A domain-agnostic, open source scientific database.
“Protein Data Bank-in-a-Box”
I'm working on a "simple" (started out that way at least) ETL app for the public library system that I work for; the target output is SQLite databases where Datasette will be the platform to interact with the data being generated/extracted. Primarily, the goal is to provide a simple report tool for staff and give me a source for nightly snapshot outputs to a datalake in parquet format.
I finally decided to promote my gitignore pattern Python library, pathspec, from v0.x to v1 after 12 years or so.
I'm thinking of reviving my Python SQL parser prototype I have half done. Or maybe resume my Mako template plugin for PyCharm.
I am working on a query cost analyzer that looks at query performance and does maths against your infrastructure cost to understand what is your dollar cost per query. The idea is that teams (when tables are tagged appropriately) will be notified with their most costly queries and hopefully makes that actionable for teams to clean up some queries.
I am unsure if there is a need for a tool like this in the market but I am becoming more and more curious around databases so this felt like a lower barrier for my product-minded engineer skills to get into.
I was working on an Arazzo[0] generator as a plugin for the Serverless Framework. This has now become a fully fledged runner. So far I have it using multiple OpenAPI source descriptions, and just about got retry rules working. Next steps will be for it to be able to reference Workflows in external Arazzo documents.
I'm still working on my Web Server Library .NET Core
I'm rewriting from scratch : https://simplew.net/v26/
I'm making a synthesizer drone generator (Webaudio API) with an audio visualizer (Three.js). It's a fibonacci sphere made of 100k particles. I send a sine wave down the spiral and it highlights the different spiral paths contained within the larger fibonacci spiral, which creates some really interesting patterns. I'm currently working on the audio component.
https://github.com/syamp/biscuit - Just a fun experimental tsdb completely written by codex.
Working on creating dashboards from google sheets, csv or json here EasyAnalytica.com
A little desktop app that lets me upload transaction csvs from my bank and figure out how much I need to split with my partner. Mostly because I always forget to charge her for utilities or flight bookings and I hate going through the bank UI. Might also expand it with some simple subscription auditing logic.
Also, a dramatic anime intro (complete with cheesy AI generated theme song and video) starring our foster kittens. It's been interesting to learn about some of the techniques needed for consistency, how to storyboard, etc.
I’m building a simple CSS redesign for Hacker News on Safari iOS as an extension: https://apps.apple.com/us/app/y-redesign-for-hacker-news/id6...
I have a few other small side projects that generally improve my day-to-day life, including a better calendar widget for shift workers and a video speed controller that floats on websites where I frequently watch videos for easy access.
My project is WithAudio and is gonna be WithAudio for a while. Its a text to speech reader. Initially I decided to generate pargraph by paragaph. But that was not a great call as users sometimes might have to wait for the whole paragraph to be ready before they can listen. Now I'm working on changing it to sentence by sentence. I think that + adding 2 new languages would take most of my January's budgeted time.
DTAP protocol.
Double TAP is lightweight testing framework where users write black box tests as rules checking output from tested "boxes". Boxes could be anything from http client, web server to messages in syslog. This universal approach allows to test anything with just dropping text rules describing system behavior in black box manner.
Working on a list of examples and by extension, a greater understanding of the appeal to Ludditism: https://docs.google.com/spreadsheets/d/1M_UjOPxpbKMYes5CcWRW...