Ask HN: What are you working on? (February 2026)

back working on my lighting desk, after a couple of years of hating it because the communications bus between the many different modules was flakey and so the whole thing wasn't fun to use. I bit the bullet last year and re-implemented everything with CAN-bus communications and it's actually fun to use now.

Current work has been improving boot time. Was nearly two minutes because of one board, and that's a long time for the lights to be out if you have to reboot during a show. I'd wanted to use buildroot to get a custom kernel that should boot much more quickly, but the buildroot learning curve was steep for me, particularly as I've no expectation of ever needing the knowledge again.

Independently but concurrently I decided I really ought to understand what all this AI stuff was about, for fear of getting left behind. That coincided with the release of opus 4.5, and holy heck has it made a difference! With a little guidance from me Claude got the buildroot environment working and the boot time down to less than 10 seconds. I've been _really_ impressed. I've had Claude write a few boring utilities that I could easily have done but Claude managed much faster and with less boredom on my part. Fortunately for my AI revolution I think I'm a better Business Analyst/writer than I am a coder, so it fits with my temperament.

I study German and decided to write app to simplify some tasks: 1) generate anki cards from text 2) extract highlighted words from paper and generate anki cards 3) collection of texts/dialogs on random topics 4) and so on... )

not sure if someone needs it, but very helpful for me )

Built my own 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.

https://riffradar.org/

https://github.com/debba/tabularis

I’m working on tabularis, a lightweight desktop database management tool designed for developers. It provides a modern interface for managing MySQL/MariaDB, PostgreSQL, and SQLite databases through a native desktop application. Built using Tauri v2 (Rust backend) and React 19 (TypeScript frontend), it offers native performance while maintaining the flexibility of web technologies.

Building PushMe: A service that sends you notifications for anything you care about.

I was wasting way to much time scrolling news. It consist of a crawling engine, de-duplication, llm creating a custom keyword filter and each event is checked against the original prompt for verification using a llm before sending it out.

Keywords suck, 381 of 388 potential matches end up like this:

"The event is about an OpenClaw Plugin Hub supply chain attack, not a data breach affecting hundreds of millions of users. The watch request specifies a massive data breach; the event does not clearly indicate that."

[https://pushme.site]

Making improvements to https://engineering.fyi/ - an aggregator for big tech blogs

Added summaries so readers can quickly see if it's worth reading the entire article e.g. https://engineering.fyi/article/towards-self-driving-codebas...

I was teaching coding to my 10yr old and we were talking about creative projects on the internet. That led to discussing the Million Dollar Homepage and why something that simple worked. He asked: could we build something similar today? That curiosity turned into moltbillboard.com — a simple public billboard, but born in the era of AI agents (inspired by the recent OpenClaw craze). It’s just an experiment..

Working on reproducible test runs to catch quality issues from LLM providers.

My main goal is not just a "the model made code, yay!" setup, but verifiable outputs that can show degradation as percentages.

i.e. have the model make something like a connect 4 engine, and then run it through a lot of tests to see how "valid" it's solution is. Then score that solution as NN/100% accurate. Then do many runs of the same test at a fixed interval.

I have ~10 tests like this so far, working on more.

Working on https://allscreenshots.com - a screenshot API for developers.

I got frustrated with existing screenshot services choking on cookie banners, rendering half-loaded pages, and serving bloated images. So my co-founder and I built one that auto-dismisses cookie consent dialogs using Playwright heuristics, serves AVIF-first from Cloudflare R2, and supports geo-distributed rendering so you can capture pages as they'd appear from different regions.

Spring Boot + React + PostgreSQL. Bootstrapped after selling a previous ecommerce SaaS.

Currently documenting the whole build in a 30-day series on the blog if anyone's into that sort of thing.

https://agentset.ai/

Open-source RAG infrastructure.Every team I talk to has the same experience: RAG works in the demo, breaks in production.

We handle ingestion through retrieval with optimizations baked in. 97.9% on HotpotQA vs 88.8% for standard RAG. Model-agnostic, 22+ file types, built-in citations, MCP server. MIT licensed.

https://github.com/agentset-ai/agentset

Building TimeCap: an app that removes the addictive sections of any social media. That means that you can check your friends' updates on Instagram without getting distracted by reels.

It is fully customizable, works on any social media, and it genuinely brought my screen time to less than 1 hour.

It is free to use for one platform, and is $29.99/yr for unlimited. And if you need the premium version but can't afford it, just send me a message and I'll sort you out.

App Store link: https://apps.apple.com/us/app/id6737515680

Have been working on three micro-saas, all built in Elixir/Phoenix:

https://feedbun.com - a browser extension that decodes food labels and recipes on any website for healthy eating, with science-backed research summaries and recommendations.

https://rizz.farm - a lead gen tool for Reddit that focuses on helping instead of selling, to build long-lasting organic traffic.

https://persumi.com - a blogging platform that turns articles into audio, and to showcase your different interests or "personas".

Maintain my blog

https://www.bboy.app

Built this (well, forked and adapted) this morning: https://github.com/leochatain/zsh-claude-code-shell

Couple of utilities for using claude code in my zsh. #? <describe command> => generates the command for you #?? <command> => explains the command for you

Packaged as an oh-my-zsh plugin so it's easy to use.

Working on SubSmith, a language learning tool to help with immersion through auto transcription, popup dictionary and Anki integration.

https://subsmith.app/

Solar-powered data center in a desert.

Fully off-grid using solar, batteries and Starlink for uplink. Focusing on AI inference at the beginning. Currently building our first prototype and testing cooling solutions.

https://solarcube.com

https://catchditto.site

I'm building Ditto — it clones websites with 100% visual accuracy and outputs a proper React app with named components and preserved structure.

The problem: you find a design you love, want to use it as a starting point, and your options are either manually recreating it or using a tool that spits out a tangled mess of divs and inline styles. CatchDitto gives you an actual codebase — clean component hierarchy, sensible naming, structure you can extend without wanting to rewrite everything first.

I'm still iterating, would love to hear what others think.

Fine tuning Gemini 2.5 flash on EVM Blockchain execution structures and using perplexity with PageRank to perform transaction anomaly detection. https://lite.chaingenius.ai/theory

[SMILES](https://en.wikipedia.org/wiki/Simplified_Molecular_Input_Lin...) or [SELFIES](https://resources.wolframcloud.com/PacletRepository/resource...), but for EVM Blockchain executions.

Finding work after a corporate restructure. Also migrating my workloads from VMs and strewn-about containers onto a Talos K8s node, so I can break the cycle of bespoke builds at home and get back to enjoying projects.

Speaking of projects, I’m roughing out a method of pulling cost data for common services (compute, storage, databases, etc) across the three major cloud providers and making recommendations as to where to put things for optimal cost; a key component of a “universal cloud” idea I’ve been kicking around since 2020 or so, where the base cloud services are abstracted away into commodities rather than bespoke products or locked-in vendors. The goal is to basically have something like Terraform that will transpose its code to the destination cloud chosen by the cost analyzer at execution, and eventually auto-migrate load as prices or needs change (e.g., a client churning early and shifting that reserved instance to another customer for a higher margin).

Write once, and trust the pricing model to deploy it where it makes the most fiscal sense. No more learning Azure/GCP/AWS for bog-standard workloads anymore.

It's still in beta but I repackaged Descent Raytracer (a remaster of Descent (1995) made by students at Breda University) to be launchable on macs with Apple Silicon (ray tracing reqs M3+).

https://github.com/rdavison/DXX-Raytracer-ar/releases/tag/ar...

Forge – A 3 MB Rust binary that coordinates multiple AI coding agents via MCP https://github.com/nxtg-ai/forge-orchestrator

Forge is an orchestration layer that sits between AI coding tools and your codebase. It's a single Rust binary (~3 MB, zero runtime deps) that runs as an MCP server over stdio. Any MCP-compatible AI tool can call it.

MIT licensed. Whitepaper with the full architecture: https://nxtg.ai/insights/forge-whitepaper

I am getting the AI Agents to build an expense tracker Telegram. I would like to have one myself and among my family members since we are heavy Telegram users. I am also using this as a way to learn more about the AI Agents (what they are good at, their limitations, etc) with (hopefully) proper guardrails, guidelines, checks, etc.

https://github.com/yelinaung/expense-bot/

https://gitlab.com/yelinaung/expense-bot/

As you may see from the git history and "contributors", it's mostly Claude and AMP making the changes. I am not entirely sold on these agents and not particularly excited by these. But I also feel that I can't afford to sit out this transition so here I am...

I launched https://rebrain.gg/ a few days ago.

It's a bit like Reddit but focused on learning. (Doom learning instead of doom scrolling)

You 1) upload a source 2) direct the kind of questions you want to be asked 3) start answering (and if you get the answers wrong, you can discuss the problem with "AI").

You can read other people's sources, questions, answers and their discussions with AI too.

And if you're learning the same thing as other people, you can join communities to share sources/questions.

It's still very early on, so I'm very interested in any feedback.

Creating an Android app of my favourite word game. Existing games are full of ads. Started coding, thanks again, thanks to AI.

https://play.google.com/store/apps/details?id=com.superfun.w...

Currently building https://auth-email.com

Many mainstream email providers have switched to require OAuth for login, but there are tons of clients and apps that don’t (or can’t) support OAuth.

Auth-Email is a secure, private relay that takes out the need to worry about OAuth: authorize mail accounts one time in our dashboard, then use an ordinary username and password for IMAP, POP, and SMTP via our server.

Chapter (https://jordanful.github.io/chapter/) - a self-hosted ebook reader (server/client) that uses a local TTS AI model for narration. Progress is stored server side, so you can switch from reading to listening anytime.

https://grandpacad.com/

Dimensionally accurate AI 3D modelling. My grandpa has a 3D printer but struggles to use any complex tools. So I am working on this chat interface to allow him to do some simple models.

So far he has triggered more than 150 generations. It’s getting better every model cycle and gives me something I enjoy working on.

So I'm building this: https://summaryworld.life

It gives you clean text summaries of YouTube videos. There are obviously other tools that do this, but I wanted something that is aligned with actual principles of learning and retention, not just quick TLDRs.

Also added a feature called Related Videos. It extracts the key themes from a video and recommends the top 3 related videos, essentially creating a small “knowledge web” of sorts around the topic. Similar to youtube recommendations, but you don't watch you click and read.

You can do YouTube search directly inside the product. When you click a video, it generates a summary. So you’re still browsing YouTube, but click turns a video into something you can actually consume. Personally found it better way to consume youtube, quicker for me to get through the content I want to consume than have those 10+ Youtube tabs sitting in my browser forever.

There’s also a public library feature I added where you can make your summary public. It’s kind of fun to see what other people are learning.

Still early, but iterating on it, scratching my own itch.

I'm going through CMU 15-721 and building wsql[1]; a GPU-accelerated query engine. It uses wgpu for cross-platform portability and implements modular pattern, heavily inspired from Sirius[2] paper.

[1]: https://github.com/fastingrat/wsql/

[2]: https://vldb.org/cidrdb/papers/2026/p12-yogatama.pdf

A substack for 80/20 life advice and behaviour change.

https://euzoia.substack.com

Full project: https://euzoia.org

Tried to be super low-tech: Notion, super.so, Spotify creators, riverside.

Now thinking of building an email-based agent for behaviour change accountability. Would love any pointers to good UX for email-based AI assistants.

Building https://typequicker.com

We’re aiming to build the best typing application; personalized to every users typing habits.

Typing is one of the most important hard skills today and yet most education systems skip it.

Most of our customers are adults who always wanted to type but can’t find the time. We make it faster to learn and improve by focusing around the user’s weak points (with our features like SmartPractice and TargetPractice)

I am working on my first multilingual word game: https://changeword.org

My previous games have all been exclusively in English, but this one also has Spanish, French, German, Dutch and Swedish. It's a take on the classic word ladder game, with golf scoring mechanics.

Still needs a bit of fine-tuning for the word lists and puzzle generation, but I think it's already pretty playable. :)

I'm working on https://videoreject.com

Video Reject is a platform for people to buy/sell physical media. Like Reverb but for physical media.

I love movies and wanted an excuse build something in that world. Its not necessarily groundbreaking, but I think it could build a nice community of folks with the same interests.

We're getting ready to launch, but in the meantime, we have a waitlist up.

This month is dropping network cable to the home offices and then adding recessed lighting in the living room, pantry, and coat closet.

Next month prep starts for finding dev work after an extended hiatus.

http://tootterhalf.com/

I've been working on this project, which lets you create interactive Valentine’s Day invitations for your special someone. You can pick from a set of templates, add your own message or photo, and share it easily.

Currently, adding CMS feature so that user can edit their info right in the website and get the link instead of them hosting themself or editing the code.

The first AFAIK customer owned self-hosting as a service company. You own the hardware, the software and your data without the hassle of needing to configure or maintain any of it. https://skarabox.com/

The technology is using my open source project https://github.com/ibizaman/selfhostblocks

Hi!

I’m working on:

- DietPDF ( https://github.com/Zigazou/dietpdf-haskell ) an intensive PDF optimizer which tries to reduce PDF size as best as it can while preserving visual quality. It relies on a lot of tricks to achieve this goal (minification, precision reduction, deletion of superfluous operations, concatenations, filter combination etc.). At the present moment, I'm trying to implement CCITT Fax encoding/decoding.

Building an AI debugger for embedded systems.

You know that feeling when your microcontroller crashes and you spend 3 hours staring at cryptic registers trying to figure out why? Yeah, I got tired of that.

So I built an MCP server that lets Claude talk directly to GDB. Now instead of manually decoding CFSR registers, I just ask "why did it crash?" and get back "division by zero at line 142 in calculate_average()".

It's pretty satisfying to watch Claude diagnose a deadlock between two RP2040 cores in 10 seconds - something that would've ruined my entire afternoon.

Just shipped v0.1.0: https://github.com/ezulabs/embeddedgdbmcp

Still working on my LPFM radio station, www.kpbj.fm. We launched our regular online stream and have secured a site for our transmitter.

The transmit site is in the Verdugo hills and will need to be off grid. We're going to need about $20k to get the entire transmit system setup (including a solar/battery backup system).

I've also been working on our web infrastructure. The site is built with Haskell and HTMX. The stream is Icecast, and stream scheduling is done using an internal schedule system in the web app and LiquidSoap (so no external tools like airtime, libretime, or azuracast).

I’m working on an open source CLI that experiments with governance at inference time for autonomous systems.

The idea is to let multiple agents propose, critique, and stake on decisions before a single action is taken, rather than letting one model silently decide. It’s model-agnostic and runs locally, with no blockchain or financial layer involved.

I’m mostly exploring whether adding explicit disagreement and cost at decision time actually improves outcomes in high-stakes or automated workflows.

https://github.com/consensus-tools/consensus-tools

I've also created an AgentSkill to interact with the cli:

https://github.com/kaicianflone/consensus-interact

ragextract.com

A document RAG API based on multimodal embeddings that's intended for data extraction. If your document workflow involves search and you're looking for ways to cut down on VLM (OCR) costs, Ragextract provides a simpler alternative (less bells & whistles) which makes sense for startups, SMEs and freelancers.

As someone who works on AI document workflows, I use Ragextract myself to reliably execute for various clients in finance, insurance and proptech. I'm currently working on marketing/messaging for the service and could do with more feedback and use-cases.

Have a new or existing project which could use something like Ragextract? Email me and if there's a fit happy to provide a demo or free subscription.

Learn more here: https://subworkflow.ai/blog/ragextract/introducing-ragextrac...

Orchestera (https://orchestera.com/) - Fully managed Apache Spark clusters in your own AWS account with no additonal compute markups, unlike EMR and Databricks.

Currently implemented the following:

- Automated scale in / scale out of nodes for Spark executors and drivers via Karpenter

- Jupyter notebook integration that works as a Spark driver for quick iteration and prototyping

- A simple JSON based IAM permissions managementent via AWS Parameter Store

Work-in-progress this month:

- Jupyterhub based Spark notebook provisioning

- Spark History Server

- Spark History Server MCP support with chat interface to support Spark pipeline debugging and diagnostics

Open to feedback and connecting. Docs at https://docs.orchestera.com/

I'm working on Skyscraper for Bluesky, which is an independent Bluesky client for iPhone, iPad, and the first Bluesky client for Apple Vision Pro.

https://apps.apple.com/us/app/skyscraper-for-bluesky/id67541...

Been a side project for a few months, as I wasn't enjoying the official client, and wanted something that fit my particular tastes (scroll position permanence, greater emphasis on hashtags, support for Shortcuts, and a more iOS-native look and feel).

Launched 3 weeks today, and already nearing 100 subscribers (nearly all annual, but month-to-month also available). Been very fun seeing users start using the app each and every day!

StoryStarling - Turn your story idea into a printed children's book

https://storystarling.com

Working on a platform where you describe a story concept and it becomes a real, illustrated picture book - professionally printed and shipped to your door.

The key difference from "personalized" book companies: this isn't template stories with a name swapped in. You bring an idea - maybe a book about a kid with a cochlear implant going to their first day of school, or a bilingual German-Turkish story about visiting grandma's village - and it generates a complete original narrative with consistent illustrations throughout.

You can upload reference photos so characters actually look like your child. Supports 30+ languages including bilingual editions on the same page.

Currently refining the showcase features and adding RTL language support.

Building pkgstore.io, a marketplace for commercial .NET NuGet packages.

https://www.pkgstore.io

NuGet.org doesn't distinguish between a hobby project and a professionally maintained library with real support. pkgstore is a curated directory and marketplace where publishers can sell NuGet packages directly, with full dotnet push/restore support, Stripe payments, and automated access control.

In open beta now, onboarding publishers. Would love feedback.

https://sept-marriane.com

An all-in-one operations workspace that starts with forms but doesn’t stop there.

teams (especially HR and ops in small/medium companies) need to launch and operate fast, but the moment data comes in, everything falls apart into spreadsheets, emails, and half-used tools. the real work starts after collecting data.

What we’re building:

Form & workflow engine, and backoffice layer on top of submissions: tasks, assignments, reminders, budgets, client/contact directories, inventories, meetings, internal knowledge base.

Operational add-ons: quizzes & assessments, certificate generation, invoicing, document repository, reusable email templates, Anomalies module.

It’s less “Typeform competitor” and more “what happens after the form”, for HR teams, marketing teams, and SMEs that want one calm, predictable system instead of a tool stack.

(we’re Morocco/EU-oriented).

Still super early (one month), shipping steadily.

I just proved that constraint solving problems can be encoded as p-adic linear regression problems[+], and that therefore we can use machine learning optimisation techniques to get exact answers.

So of course no journal or conference is in the least bit interested, and I'm now reformatting it for another obscure low-tier journal that no-one will ever read.

Otherwise:

- automating the translation of a Byzantine Greek work that has never been translated into English before. https://stephanos.symmachus.org

- also preparing evidence for a case against the university I sometimes work for.

[+] Linear regression, but instead of minimising the Euclidean distance, minimise the p-adic distance - https://en.wikipedia.org/wiki/P-adic_valuation

Applications for Linux that I always wanted but could never quite find the one that works how I think it should.

traymd: A system tray notes application that supports basic live input of markdown. https://github.com/rabfulton/TrayMD

reelvault: A local film browser and launcher. https://github.com/rabfulton/ReelVault

preditor: A simple image viewer that shows each image in the center of the screen in a window sized for that image with some basic editing functions built in. https://github.com/rabfulton/preditor

I’ve just published the first public release of a new open source project Shovel.js, replacing tools like Express, Fastify, Next.js, Vite. It’s a full-stack/meta server framework which implements the full Service Worker specification but in Node, Bun, Cloudflare. It leans into using web standards to do things like accessing the filesystem, reading cookies, create client-side bundles rather than inventing new APIs. You can read about the process of making Shovel with AI in the introductory blog post.

https://shovel.js.org/blog/introducing-shovel/

https://github.com/bikeshaving/shovel

Most parents I've met understand the internet can be a dangerous place for children but aren't sure what to do about it. Some avoid it altogether, but most give up and resign themselves to whatever happens. What I wanted most was just to have some visibility into what my kids were experiencing, so I built LivingRoom App for iPhone and iPad. It sends occasional screenshots to parents. My hope is that when we shine a light on the online world, we will be free to use the internet as a tool for learning, creativity and connection. https://livingroomapp.com/