alt Hacker NewsAsk HN: What are you working on? (May 2026)
What are you working on? Any new ideas that you're thinking about?
Comments
Working on https://tapitalee.com
- Deploy containerized apps to your own AWS account with minimal config!
- CLI tool with instant console sessions
- Set up SQL/Redis instantly with Heroku-like add-ons.
- For enterprise: Autoscaling, preview apps, audit trail, release approvals.A model framework for an in house suite of models.
From dataset harvest, to training intricacies on CUDA/ROCm to fun HIP kernels. Full circle to inference testing, building it around consumer hardware(the challenge). Using this as a "how it works" deep dive, allowing me to learn more about the how, more than endless papers will. It's a MoE and I'm slowly running a human loop, research, build, correct, research.
http://autonoma.ca/calculators/rocket/antimatter/
Given a distance, an allowable time to reach that distance, a payload to send, and an expected exhaust velocity, how would you calculate the time required to convert energy into antimatter fuel and how much antimatter needed to arrive at the destination (starting from the Moon)?
There are a few side calculations, such as the size of the radiator, estimated footprint of the fusion reactor itself, and how much metamaterial is needed. This is to help figure out timelines for a sci-fi novel, so ballpark answers are completely fine.
The calculations yield what appear to be values around the correct order of magnitude. Would be delighted to have insights, comments, and corrections.
As a data-obsessed golfer trying to get to single digits, I need a tracking app that picks up where Arccos leaves off. So I'm building one: https://shortgamewiz.com (still a bit WIP).
After a few rounds of using it, I already know a few things I didn't before: I suck at right-to-left breaking putts, I baby uphill putts too much, and getting out of bunkers consistently is not good enough if I can't sink the occasional save. So I know what to practice now.
I track my learning and schedule repetitions in google sheets. But Google sheets sucks on the phone. So I built a dumb frontend reading off of my (public) google sheet which just has 4 columns for links, title, dates and wait times, plus a formula. Webapp pulls the sheet as csv, renders as color coded lists and a couple charts. Chart shows what's due this week on a 15 week timeline. This is the simplest luddite version I could come up with. I don't have a way to share this with others except sharing the source. Not introducing complexity from auth, storage, managing updates from the app, etc.
Been writing in my blog every day, reading more, created a poker equity calculator, and working on a city wide project where I document attractions, restaurants, and stays I've experienced in my city (very early stages).
Website: https://arkvis.com
Poker Equity Calculator: https://github.com/lodenrogue/poker-equity-calculator-web
Davao Explorer: https://github.com/lodenrogue/davao-explorer
Reading Summaries: https://github.com/lodenrogue/reading-summaries
I also created a couple of chrome extensions:
HN Dracula Dark Theme: https://github.com/lodenrogue/hackernews-dracula-theme-chrom...
Regex Search Chrome Extension: https://github.com/lodenrogue/regex-search-chrome-extension
Created a small command line util to get earthquake data in the Philippines:
Philquakes: https://github.com/lodenrogue/philquakes
I'm working on a little local first review tool called Review (though I sometimes refer to it as differ since that's it's original name) - you can see screenshots here https://x.com/rhyslikepb/status/2053149881104265599?s=20
The idea was borne out of wanting to use the review tools that you get on existing sites like GitHub, without having to push and start bloating PR lists. You'll be able to leave yourself comments and code suggestions after review, which you can then pull out in a Markdown file to feed back to your coding agent (or anything else for that matter).
I'm also trying to include some optional (very optional) AI extras where you can use your own keys, and then get a tour of what you've changed and a quick overview of the changes.
I recently switched to developing VST audio plugins and I'm loving it. Already done 3 [0][1][2] and I want to keep doing this if I manage to generate some income from it. I develop them in Typescript and then convert them to C++ with Webview, this way I have a web demo of the plugin that is almost identical to the one you get for the DAW.
[0]: https://technokick.com/ (Techno Kick synth)
[1]: https://riviera-demo.surge.sh/ (Reverb effect)
[2]: https://ya3.surge.sh/ (TB-303 synth clone)
Paste Redactor. It redacts Personable Identifiable Information (PII) from your clipboard when you copy and paste text. It uses a custom trained local AI model so that your PII never leaves your device. That is what it does now. Currently working on it to make it work for agents as a privacy protection layer. The idea being that the most powerful AI models live in the cloud but need access to your local files to be useful. We instead want everything to go though a local protection layer before it is sent to the cloud possibly with labels and then reconstructed locally when the cloud sends back its results. Kind of like a Adblocker but for agents and private data instead.
I am working on a framework that lets you easily create tools inside the Django Admin - https://djangocontrolroom.com
I've published several panels under this banner already (tools for redis, caches, celery, etc.); I am currently working on a base library layer for tools to inherit from and to make it easier to create new tools.
Essentially, the point of all of this is to make it so that you don't need so many external services; Instead, DCR provides self hosted alternatives. This in turn makes it a lot easier to build and productionalize something using Django.
Reception has been decent so far and I estimate several thousand current adopters (Its hard to estimate based on download numbers alone.) For May I will finalize a common design language, further formalize the plugin system and how it works, and likely release a new panel.
I have been working on two opensource tools:
https://dhuan.github.io/mock/latest/examples.html Command line utility that lets you build APIs with just one command.
https://github.com/dhuan/dop JSON/YAML manipulation with AWK style approach.
I've been working on a pure Clojure implementation of WebRTC Data Channels (SCTP over DTLS over UDP). The library provides a minimal, dependency-free (except for Clojure itself) way to establish peer-to-peer data channels on the JVM.
I've always wanted this and have used it to experiment with Gemini's cloud agent Google Jules.
I made https://poemd.dev/ as an online markdown scratchpad that supports GitHub Flavoured Markdown and stores all data in the URL. This means there are no accounts to work with and everything is basically stored in bookmarks if you choose to.
The persistence model makes documents somewhat sharable, but I do find Open Graph previews to be mixed. In Messenger it renders the whole URL, which is quite long due to encoding, and that kills the conversation view.
I am trying to better my understanding of Agentic coding tools and Tool using Agents by building my own, without relying on an Agent that writes said Agent/Harness
I believe writing my own "Toy Harness" is a good way to learn and understand these tools.
Other than that, I did plant my tomatoes today.
I am working on a research institute for East Africa, https://maiyoinstitute.org/. I want to tackle the dire lack of environmental data, by using 1. low cost hardware 2. Artificial Intelligence 3. Long term horizon. The problem set is huge, but I am focusing on low cost sensors for Air and Water data collection plus bioacoustics for now.
Working on trying to give some guidance on solar panels.
Plug in solar became legal here in the UK
Still sussing it out but started shipping something
Finding the pitch direction of the roof is kinda hard
Uses data from the house to try and get a rating
Working on a bread recipe community where you can share and iterate on bread recipes. It's out of personal interest to be able to record my bread recipes and thought it might be interesting for others too.
However, I worked on it for the past ~5 years on and off (well, mostly off) and rewrote it too many times. Now finally close to releasing, bought a domain and setting up all the last remaining things.
Still working on my SQL canvas: https://kavla.dev/
I started with this last summer. Usually I get tired of an idea, but this one is just an endless pit of things to try out.
Currently seeing how we can get an analytics agent working on the canvas. Video here: https://x.com/i/status/2053410747137266070
We're working on Webhound - budget controlled long-running deep research. You set a budget and Webhound will use that much in compute/LLM tokens to research your prompt, with built in verification cycles and optional added verification budget. Every claim is cited with evidence and a direct link to the tool calls that produced the claim
The goal is to build a deep research product for actual researchers, since we believe that it is an extremely powerful product that is still nascent but has enormous potential - which we've already seen with some early users.
GBP (Google Business Profile) is getting extremely popular for B2C business marketing - e.g., https://trends.google.com/trends/explore?date=today%205-y&q=...
We just received the API usage approval from Google, and I'm integrating GBP to https://pinpost.io this week (our reliability first social media management tool)
I've been working on Mesaphore, an Excel-like spreadsheet app[0] backed by a Parquet-based file format. The premise is when Excel starts off as the starting point, then over time becomes a data exchange format between systems, and eventually becomes a bottleneck for the system. You still want to provide your users something Excel-like but also want to address the limits of Excel[1].
Majority of code (almost 70%) is generated by Gemini Pro and is extremely ugly. Due to a recent eye injury, I've not been able to code as much as I want, so I'm delegating many things to Gemini. Eventually, as my health improves, I plan to rewrite the entire thing.
[0]: https://codeberg.org/naiyer/mesaphore
[1]: https://support.microsoft.com/en-us/office/excel-specificati...
1) “AI harness plugin build system” to help improve reliability of and increase compatibility across the fragmented AI coding harness plugin ecosystem.
https://github.com/jondwillis/jacq
2) Claude code plugin based on some ideas found in https://www.anthropic.com/research/emotion-concepts-function The main idea is to add hooks that inject “baselines” under some conditions to counteract certain “emotions” that can cause subtle misaligned behavior in agents
https://github.com/jondwillis/functional-emotions
3) Final Fantasy XI custom client remaster in Bevy/Rust alongside an MCP integration that aims to allow agents to play autonomously on private servers à la “Claude plays Pokemon”
Contact: https://jonwillis.dev
Closing up work on my modular, hobbyist, analog computer. (Finishing up the manual—the hardware is already a wrap).
Something I can finally enjoy: just playing with it. I tediously wired up a pair of pendulum simulations to drive an XY oscilloscope—got a nice Lissajous curve.
But now I want to double it to four pendulums. Each axis (still just X and Y) to be driven by the sum of a pair of pendulums. With them out of phase, the curves appear to sometimes collapse but then suddenly explode again…
(Love to eventually hook it up to an actual plotter.)
Building a custom feed for Bluesky which uses collaborative filtering over the likes data: https://foryou.club
How the algorithm works: it finds people who liked the same posts as you, and shows you what else they’ve liked recently.
Launched the feed a little over a year ago and it has become the most liked feed.
I'm slowly but surely working on a first update to my Android app, Tunemark (https://tunemark.app), which I released a while ago. Tunemark lets you add bookmarks to moments in songs so that it is easy to jump back to them. It is really convenient when practicing dancing or music and you need to constantly reset back to specific parts of songs. Unlike most DJ-type apps that could serve similar use cases, Tunemark works with most music apps, including streaming services.
I have new features such as sharing bookmarks and possibly BPM detection planned but also some quality of life changes like better UI scalability for different size screens/split screen use.
Been working on and off on a Spotify recommendation egnine after getting tired of Spotify’s repetitive recommendations.
You get to choose the genres you're interested in, and it creates playlists from the music in your library. They get updated every day - think a better, curated by you version of the Daily Mixes. You can add some advanced filters as well, if you really want to customise what music you'll get.
It works best if you follow a good amount of artists. Optionally you can get recommendations from artists that belong to playlists you follow or you've created. If you don't follow much or any artists, then you should enable that in order for the service to be useful, as right now that's the only pools of artists the recommendations are based on.
I have been working for some time on a budget body/facial mocap solution with Unity. Mocap is hard, and what exists is locked behind subscriptions or is just very expensive.
With Unity I'm trying to bundle a bunch of different free, cheap or open source solutions together. For facial, that includes a custom converter from the output of Deadface (based on Mediapipe) with ARKit blendshapes, and also eye movement. For body it's a custom hook to SlimeVR that allows you to mocap with cheap-ish IMU-based DIY trackers, and all that on top of a custom made (not free but open source) physics rig solution that gives you accurate rigid body real time collision, saving on cleanup work.
It's being going really nice despite being an unusual workflow. Hope to release it as a plugin for a in-development sandbox game in the near future. Mocap and animation has been my passion long before i started with tech stuff, and finally I'm able to pursue it.
Instant linux boxes via ssh that suspend on disconnect. Checkout https://shellbox.dev
The idea is to have "real" linux, exposing ipv6, supporting nested virtualization, docker, etc.
My wife and I continue to work on Uruky, a simpler and cheaper Kagi alternative, based in the EU [1].
Since last month we’ve stabilized the search UI/UX and have 5 search providers you can choose from and sort as you prefer.
We entered May with over 50 paying customers and have recently launched Uruky Site Search [2] (for website owners, this effectively is our own search index and crawler, which we’ll be bringing into Uruky soon as another search provider option)!
Customers really enjoy the simple UI (search doesn’t require JavaScript) and search personalization (from choosing the providers to the domain boosting and exclusion). We also have hashbangs (like "!g", "!d", or “!e”) when something doesn’t quite give you what you’d expect, though.
You can see the main differences between Kagi, DuckDuckGo, Ecosia, etc. and Uruky in the footer (right side), but one huge difference is that with Uruky, after being a paying customer for 12 months, you get a copy of the source code!
Our main challenge right now is outreach because we want to do it ethically, and it’s hard to find communities or places to sponsor which are privacy-focused and don’t require €5k+ deals. Ideas are welcome! We’ve been sponsoring a project per month (Qubes OS, The Tor Project, and Hister so far), with our limited budget of ~$100 / month.
Because of bots and abuse there isn’t a free trial easily available, but if you’re a human and you’d like to try it for a week for free, reach out with your account number and we’ll set that up!
Thanks.
Employee benefit plan analytics. Had a huge dataset long ago as a consultant to the industry and finally vibecoded up a decent frontend. All public data but if you know the data there is a bunch of analytics you can do. Just about to launch and do some marketing in a few weeks, so saw this and thought I'd throw it in!
I'm working on a tool to let sfotware developers write well-formatted ebooks and printed novels: https://frequal.com/epublish/
Example book here: https://www.amazon.com/dp/B0GYCZJVGX
An EU replacement for PagerDuty, focusing on the absolute basics - SSO as the minimum even on free, no AI driven workflows, overviews etc. but may include ML/AI driven insights in future since that’s the way the world seems to be going.
I'm building aiki.sh so Claude and Codex can work together without babysitting
I'm working on JRECC, a Java remotely executing caching compiler.
It's designed to integrate with Maven projects, to bring in the benefits of tools like Gradle and Bazel, where local and remote builds and tests share the same cache, and builds and tests are distributed over many machines. Cache hits greatly speed up large project builds, while also making it more reliable, since you're not potentially getting flaky test failures in your otherwise identical builds.
I continue to write. Unsure the end goal but in my recent dispatch I learned to get out of my way, stop intellectualizing life and let it be.
muster and muster-pattern-library. (https://github.com/azide0x37/muster)
an agentic coding scaffold/framework you can reference when building out your next random raspi project. prefer to build around systemd units first; make an idempotent installer script, then put as little as possible custom coding around that.
`impl muster` comes down to: /build out this tool wiring together `patterns` like: C3.dropfolder-trigger; R2.device-binding; C4.lazy-resource-gate
or composite patterns like:
T2R4.device-triggered-conveyor "Bind a physical device event to a bounded ingest job that waits for hot-storage capacity, proves cold-storage capability, stages local work, and hands output to a hot/cold conveyor."
I need to back up a couple hundred DVDs, so with muster I get out:
dvd-ingester T2R4.device-triggered-conveyor
Architecture DVD media becomes ready -> udev rule adds SYSTEMD_WANTS=dvd-rip@%k.service -> systemd runs /opt/dvd-ingester/current/bin/dvd-rip-one /dev/%I --apply -> dvd-rip-one proves DEST_DIR and waits for HOT_DIR capacity -> completed rip moves to HOT_DIR/<run-id> -> dvd-publish-one.timer drains HOT_DIR to DEST_DIR -> publish writes DEST_DIR/.incoming-<run-id> and atomically renames final output
Pipelined; ejects after rip completed. Monitors local disk capacity, retries after NAS comes back online; resumes after random reboot; etc.
https://brettkoonce.github.io/lean4-mlir/blueprint/
https://github.com/brettkoonce/lean4-mlir
I (w/ Claude) have built a framework for writing neural networks in Lean 4 that compiles to StableHLO MLIR and runs on GPU via IREE.
It's been three months now of building Crit - https://crit.md - local first, open source tool for reviewing markdown and code output from your favourite AI agent.
It's inspired by GitHub PR review workflow, only with quick iterations and local.
It's been great! I found some dedicated users, dogfooding it every day with Claude and starting to get more contributions from the little community. We just got accepted into Homebrew core which was my target.
I'm expanding the team features now as I've got a few users keen to get the sharing service deployed in their private networks!
I've been iterating on the type system of a somewhat-kooky programming language I'm developing: https://github.com/mkantor/please-lang-prototype
I've been going through Nora Sandler's Writing a C Compiler book and writing a compiler in Python. I'm excited to start the chapters on optimization - those seem like the most fun algorithm problems.
I recommend the book. It certainly isn't easy (maybe 3x harder than Crafting Interpreters), but I've learned a ton (eg how to deal with operations on different sizes of types, or the trick of using pseudoregisters to avoid having to figure out registers up front).
I’m working on a ground-up implementation of RADIUS with everything running on stateless compute. It’s a beast with many problems to solve but I have EAP-TLS, TTLS and PEAP all working. I’d love to connect with folks interested in this kind of thing.
A small, generic Go library for retrying fallible operations with exponential backoff and pluggable jitter strategies. https://github.com/nodivbyzero/try
Building a small voice journaling app, hoping to ship in the next few days https://turquoisehexagon.co.uk/anima
TestFlight link, good for 10 users: https://testflight.apple.com/join/9VREtXzq
I'm building DB Pro, a modern, beautiful, and now fully self-hosted database client for desktop and web.
Just launched Studio, which is the self-hosted version of DB Pro.
I also keep a devlog. #9 was just published to YouTube.
Self-Host Your Own Database Client | DB Pro Devlog #9 https://youtu.be/MJvSrJGtk70
I am building Akariq which provides data eSIM data plans to 185+ countries and regions across the world. I am 2-3x cheaper than big brands in the same space. I also prefer local data routing i.e. I don't route traffic from US or EU to Hong Kong, they stay in your country / region.
I'm making a surf forecasting site a la surfline.com, I started mostly to have an API to use for my tidbyt, but I figured I might as well make it a full thing and built my own features! It's on quickswell.com but it's only Socal at this time (fewer spots to compute)
Menu bar app that reduces your Claude Code token costs by ~50% so you get 2x more usage out of your plan.
People seem to like it so far :-)
I’m building a UI design app similar to Figma or Sketch, but with a few differences:
1. Responsive artboards and flex-like layout engine
2. Deep support for design tokens
3. HTML/CSS previews and export
4. Multiplayer AI and human collaboration. Agents can connect to documents and collaborate like any other user.
Built in Swift and cross platform Mac, iPad and iPhone.
I’m designing and building the UI and implementing the underlying features with Codex. So far it’s going surprisingly well.
Main project is a deterministic .NET runtime (https://github.com/Smaug123/WoofWare.PawPrint). Today I upgraded it to net10, which has naturally caused dozens of regressions which Claude is beavering away at.
Side project is my own agent harness, https://github.com/Smaug123/writ , which is being built sandbox-first and with Nix as a first-class citizen. Obviously everyone has to write their own agent harness as a rite of passage.
AI basketball community, using computer vision to get highlights and stats http://ballers.gg