What are you working on? Any new ideas that you're thinking about?
Just tried out Opus 4.6 to make a ground-up new version of a perennial side project: static site podcast player.
For this one I focused on loading speed and reducing interaction with repo. So it processes the images (converting to webp) and loads the feed list from a Gist. Also used the "frontend-design" skill. From brief to ready-to-use took about a couple hours.
I have been working on chemistry arena designed to benchmark the current SoTA LLM's on drug discovery tasks. My plan was to then focus on getting this annotated to sell this data for post training of scientific reasoning models. https://github.com/deepakorani/chemistry-arena.
Also trying to find a co-founder who I can work on projects and solve problems faster to be honest.
I just released Configmesh this week. It's a macOS app (with CLI companion) for e2e encrypted syncing and backing up of dotfiles and application configurations. You can sync for example stuff from ~/.config/, Application Support, *.plists, and so on, and add config sync to apps that don't support it natively
Fresh off the press
Hi HN,
I’m working on an API that provides heuristic signals about whether a piece of text is likely LLM-generated.
The system uses an ensemble of techniques, and on my internal evaluation set (details below) it achieves ~99.8% accuracy. I’m fully aware that general-purpose AI text detection is hard and adversarial, so this is meant as a probabilistic signal rather than a definitive classifier.
There’s a simple demo site: https://checktica.com
And public API docs (no API keys required): https://api.checktica.com/v1/docs
I’d really appreciate critical feedback, especially on:
- Failure modes you’ve seen in similar systems
- Whether this framing makes sense at all
- Where such a tool might actually be useful (or useless)
Happy to answer technical questions.
I'm still working on my Web Server Library .NET Core, the rewrite is almost finished (beta version)
With LLMs, the cost structure is credits/tokens all the way down. This month, I built a tool to help you price your service credits or plans at https://aitokenpricing.com.
I also did a write up of PRDs and using Railway to build apps on my blog at https://aftercompsci.com.
Personally:
- The Laravel Artisan Cheatsheet - https://artisan.page
- Cachet, the open source status page system - https://cachethq.io
Professionally:
- Laravel Cloud's "Private Cloud" offering - https://cloud.laravel.com/enterprise
Simplified agent task orchestrator named Kiln:
Uses your local Claude Code as the agent and GitHub as its UI, things you already have. Open source, MIT License.
You move cards across kanban columns (Backlog -> Research -> Plan -> Implement) and Kiln runs Claude locally, opens PRs, and keeps everything tracked in GitHub.
I finally got to rework my html form forwarder[1]. I realized how cheap SES is and thought I might as well make the service free.
I hope open sourcing it could be a way for me to get some mentorship from more experienced devs as well. Unfortunately, my work doesn't really do code reviews so I feel like I am not improving much on that front.
I've been working through the pwn.college curriculum, as well as building https://stoacentral.com. It's meant to be a place to chill, discuss philosophy (Stoicism), and improve yourself. I built it after becoming annoyed with Facebook groups, really.
If you're interested in Stoicism, feel free to join and start some discussions.
An on-call runbook execution engine - being able to take plain text runbook steps like - look at logs for $service, check for dependency failures - look at so-and-so dashboard.
and execute them when an alert fires.
We are at https://www.relvy.ai
I build an AI automation that finds cool things to do in my city, which is stockholm and updates the website every night along with updating the website with the events it also automatically publishes top 5 events to instagram. I use gemini nano banana model to create event posters for instagram
link to site -
Working on new code review tooling specifically for reviewing your own branches/commits when you use an "AI Agent" to assist with writing code. It seems all of the tools people are building in this space attempt to automate away the review, but I want better tools for reviewing (and tracking tech debt) in the code I just generated locally. Will publish here soon
I built https://bookcall.io. out of the annoyance existing scheduling platforms have. It's nieched towards solopreneurs, coaches, consultants, and similar that just need it to work and help them sell themselves. Also includes video calls. Would love to have feedback/pushback on the idea. But I happily use it for a while now.
https://github.com/ScivizLabVienna/HandsomeTello
An opensource iot drone for less than thirty dollars.
I'm working on Memex: https://github.com/joenewbry/memex
It's inspired by a moment a few years ago when I realized I had no history of what I had worked on in the past.
It let's you quickly get answers to questions like:
- What did I work on last week? - What was that one hacker new post about compiler optimization that I forgot to bookmark last week?
And it has MCP support so it plays well with Claude.
I've used it recently to target specific job applications by adjusting my resume based on what the job application is looking for and what I've worked on in the past... Claude one shots this (because it has context from Memex). And it feels amazing!
Also, the name Memex comes from Vannevar Bush's 1945 article "As We May Think". He originally thought of as a device in which individuals would compress and store all of their books, records, and communications, "mechanized so that it may be consulted with exceeding speed and flexibility". The individual was supposed to use the memex as an automatic personal filing system, making the memex "an enlarged intimate supplement to his memory".
And he created a word - memex - which is a portmanteau of "memory" and "index".
The wikipeida entry here has more information: https://en.wikipedia.org/wiki/Memex
It's been a slow start for me. But now, between the cli interface and the MCP connection with Claude I find that I'm starting to use it instead of:
- Bookmarks - Lists / Bug Tracking
And even more exciting it's unlocking capabilities that I didn't have before:
- Can ask Claude to review the last week of work and remind me of things that I might still need to do - Can prevent randomization when someone asks me how to configure a server that I set up a month ago. Now I just ask claude and it checks in Memex. And I can send over a nice .md file
I've been working on a low-code CRUD backend for AI agents to use to build software. To significantly reduce the complexity of deployment, access control, maintenance, devops, etc... Reducing the surface area for hallucinations and bugs when building complex apps.
Remixify[1]. What I mean to do is target DJs and people who love to own their playlist curating process. We aim to help people find remixes to their favourite Spotify playlists. Alt versions, club mixes, remixed versions, whatever. Come build your new experience.
https://github.com/askmanu/acorn
A straightforward and simple AI agent framework. It puts a lot of emphasis on the loop and the steps in that loop. You can change in real time the model, the temperature, the tools, the history. You're also able to spin-off work on a branch and then add the result of that work on the main branch. Still early but developing very fast.
I replaced MyFitnessPal: https://macros.jlcarveth.dev
Uses vision models to allow for automatic scanning of nutrition tables. Supports UPCs. Still thinking of a good name.
I also made an RPG game via a discord bot: https://questforge.ca
A voice agent calls you on your phone, and you talk with them for about 4-5 min per day to practice language immersion. Using the Twilio API and the OpenAI model.
There are issues with interrupting the agent and the high costs of calling non-US numbers. I don't think it will be a commercial business, but it's fun to play with the technology.
I am currently not really working on anything major, due to time constraints.
However had, on my semi-immediate todo list in the future are:
- improvements to the scripts I use to compile software from source; I want to be able to build a complete LFS/BLFS without any interaction (I know there is AFLS but I don't like the format or restrictions; I literally want to be able to do everything here via actionable scripts at every moment in time, including using git sources rather than old releases)
- continue on the unified widgets project (e. g. use oldschool GUIs but also for the web "GUIs"). Describe once, run anywhere, in any programming language. This is obviously way too much for a single person, so I mostly want to get the foundation right, prototype primarily in ruby, then add python and java to it. The "end format" should be for normal people, e. g. they should be able to describe everything in one format, and then have that be the basis for every GUI.
- continue working with regard to bioinformatics, also with a focus on normal people (non-tech savvy people). Most bioinformatics software was written by math-heavy tech-centric people, which makes sense. It's not trivial to work with that (we have AI to work around this to some extent, but I feel that many people who use AI don't understand the underlying components; I kind of want something like a Linux from Scratch for all bioinformatics-centric software. Like not just use it but full and useful explanations that are not too technical, but also not too overly long.)
Hopefully gem.coop becomes a viable alternative to rubygems.org - I hate the corporate identity rubygems.org adopted (and you can see the fall-out in other areas, e. g. Heroku dying right now, which kind of means ruby will die too, if all use cases are eroded despite the pro-corporate focus RubyCentral embraced). Unfortunately things seem to become worse in general - I was hoping LadyBird would be a real competitor. The moment you make any statement that they perceive as "criticism" is the moment their code of conduct kicks in and locks you out. And that's not even after a first release; imagine how they will operate once people come in with complaints about LadyBird having problems.
The world wide web used to be a LOT more open in the past.
This weekend I've been going through a bunch of stuff with A2A, building little samples and just getting my head around it. Threw together this repo[1] with a bunch of the stuff I'm doing, if anybody else is interested.
Also, watching a bunch of videos and reading docs on OpenClaw. I had thought I'd do an install of it sometime this weekend, but I don't know if I'll get to that at this point or not.
And lastly, messing with Spring AI[2]. I wanted to get a local build of that going so I can dig into the bowels of it and hack on it a bit. So I got that repo cloned and ran a quick build, and now I plan to start exploring the codebase.
I'm enjoying building a website with solitaire and puzzle games.
I am currently rewriting the engine for the Nth time and plan to add 400 games to the platform in the coming months, as well as social features such as daily challenges, awards and leaderboards.
My ambition is to make this project the largest collection of free modern solitaire games available for all kinds of devices.
I developed my first macOS app, an API client built purely with SwiftUI/AppKit. It has themes like Nord, Monokai and 8 others. Been using it for my own personal use for the past 6 months.
Making rent as an open source developer.
Shamelessly attracting new monthly sponsors and people willing to buy me the occasional pizza with my crappy HTML skills.
A dashboard to spin up web servers, proxies, remote server monitoring, remote terminal, and docker containers in milliseconds.
https://github.com/prettydiff/aphorio
Screenshots: https://prettydiff.github.io/aphorio/screenshots/index.html
Learning cribbage, my family has been learning cribbage and we are leaning hard on cribbage scoring cheat sheets, but haven't found a great one online. So I put together https://cribscore.linsomniac.com/
A developer platform for AI image generation that includes observability, with fine-tuned vision models as a judge to monitor production traffic. (Still working on the last part.)
We also have a model arena and showdown page that ranks models by task, so you can find the best model for e.g. making infographics: https://lumenfall.ai/leaderboard
We just launched the MVP. Tech stack is Rails for the dashboard and Cloudflare Workers (Typescript / Hono) for the gateway.
https://github.com/ntk148v/clicklens
It's a modern, powerful, and user-friendly web interface for managing and monitoring ClickHouse databases. It provides a comprehensive suite of tools for developers, analysts, and administrators to interact with their ClickHouse clusters efficiently.
- Effective and Native RBAC: Use ClickHouse grants to control the data access and UI permissions. - Discover - Flexible, Kibana-like data exploration for any table granted access. - SQL Console, Monitoring and Logging, ... all in just one place.
Keep working on a GUI calendar for Linux which would CLI users happy to use https://semana.alexmarkov.xyz/. Just recently I've shaped the architecture a bit.
I'm currently working on Tresor AI (https://trytresor.com) - Al confidential AI workspace for professionals (including LLMs, documents, web search, integrations, etc). We use end to end encryption plus hardware sealed AI processing, so your data is never visible externally, even Tresor can't see it.
I'm still working on basi[0], a Playwright alternative syntax/tool. I am curious about using LightPanda as an optional headless browser for it and wrote about it here[1]
[0]: https://github.com/zikani03/basi [1]: https://code.zikani.me/using-the-zig-built-lightpanda-browse...
Professionally, I'm currently working on a touchscreen interface for a medical device warming cabinet. But in my spare time I'm learning to use the micro-mill I recently purchased, by creating a cutting tool for an electric hand drill that will make flat holes in wood that are cut to exact depth and centered correctly. This is for preparing worn out pivot holes in 19th century wooden works clocks to insert bushes.
I'm building CatchIntent (https://CatchIntent.com). The goal is to help turning social conversations into qualified warm leads.
Started with it because I was struggling with finding relevant conversations about my first app where people are exactly asking for what I'm selling, only that I was missing those conversations and people. Build a POC, tested for myself and started getting good leads, so I converted it into my second app.
A fun quiz for identifying AI content on Wikipedia. Decided a game format is more interesting than markdown.
Working on pg-fs, a postgres backed file system abstraction for ai agents. So I agents can be given their familiar file primitives, without
https://github.com/DumbMachine/pg-fs
A version of it powers my local rubber duck thoughts and voice note store. Like an explicit chatgpt memory store, helps with information fomo cause I know finding the needle in haystack would be easy.
Last time this was asked I was working on this
https://github.com/fmstephe/simd_explorer
A little TUI app for interactively running different SIMD instructions and seeing the outputs.
Since then I have completed the tool for AVX/2. At this stage that's as far as I intend to go.
It's potentially valuable as an interactive quick reference guide for SIMD instructions.
It works on Windows, Linux and with the right environment variables it will successfully pretend to be AMD64 running on an Apple M chip.
Arm NEON instructions are not supported at all, currently Go's assembler does not include these instructions directly, so I didn't attempt to build for them. Maybe one day.
Next up, learn Zig - be happy.
Implementing a hobby HDL for designing circuits in Wireworld and other Cellular Automata. The eventual goal is to create a larger Wireworld computer than the original (https://www.quinapalus.com/wi-index.html). If this project actually ends up working, I may attempt to optimise some large Conway's Game of Life designs. Currently I'm at the stage of rewriting the language's solver.
WIP language spec: https://gist.github.com/Heathcorp/13fcd206fdc38ca6ce001f32ef...
Writing the compiler/solver in Rust with no AI assistance because this is a learning project.
Finally integrating Stripe! Been working on open source mobile app and ad analysis for awhile but didn't have a good flow for people to pay me. After getting 3 emails in the past month about it, and with plenty of pressure from my wife, it's definitely time.
I built a time tracking app and control panel for ourselves at our company out of frustration (it was very basic, for compliance with spanish law) and eventually we just fixed so much stuff and added so many features that we just released it as a product https://workstamp.eu
We need to reduce the entry barrier (it's meant for companies so it needs explicit registration) so anyone can use it as a proper SaaS but so far we already have a couple clients :D
Two free dmarc tools: * dmarc domain scanner https://dmarcdefender.io/tools/domain-check * dmarc xml analyzer: https://dmarcdefender.io/tools/xml-analyzer
A Postman-like MQTT platform with variables and custom MQTT topic buttons.
https://github.com/alsoftbv/topic-lab
It's a Tauri-based app so it has small binaries and supports MacOS, Linux, and Windows.
No screenshots yet, couldn't find the time for marketing work. I'm building features as I am using them. I wanted my colleagues to give it a go first before sharing to the public, but I believe it's already valuable as-is.
I built a website where kids can practice reading comprehension and learn new words while staying up to date with the latest global news each day. I originally built it for my own kids to help them maintain their reading skills while we travel. They loved it—and even suggested adding gamified features like streaks and badges!
I am working on building https://startupforstartups.com/ , single tool to manage digital presence for a small business. I have been working on it for a couple of months now.
A social bookmarking site: https://fyp3.com/
Kinda like HN meets Pocket.
It includes a Chrome extension to easily tag, save & share pages.
Currently the front page is all the pages I find interesting (AI/Startup related).
Would love any feedback or feature requests!
For the past 2ish months, I’ve been working on Lattice, my internal engine for my multi-tenant blogging system. Take a look at the code [1] and the live site [2]
[1]: https://github.com/AutumnsGrove/GroveEngine [2]: https://grove.place
Improving seccomp and landlock intergration into https://ryelang.org, improving tooling for making single executable files from rye projects, experimenting with reactive, declarative TUI library.
Building https://lenzy.ai - helping conversational AI products (think Lovable or Cursor) reduce churn and prioritize product improvements by analyzing their user's chats. I started about 4 months ago, made my 2 paying customers happy. Now trying to onboard more and more companies!