Show HN: Activiews – A privacy-first fitness alternative for Apple users Hi HN, I built Activiews as a fitness alternative for Apple users who want a simpler, more private, and more visual way to view their workouts, with a heavy focus on maps. The app reads from Apple Health and Watch data without requiring an account or sending anything to the cloud. I built it because I wanted a fitness app that: - Has no social network features - Does not store my data on their servers - Doesn't require an account or login - Stays lightweight and free of bloat - Is simple to use, with a focus on customizable workout visuals Core features: - Customizable workout maps and shareable cards - Flyover route animations - Calendar heatmap showing activity over time - Offline-first, data stays on device, no login/account needed - Dark/light themes, map styles, unit system, and accent color customizations - Reads data directly from Apple Watch - Localized in 13 languages - €0.99/month or €24.99 lifetime It’s built in Swift & SwiftUI using native APIs and focused on performance and privacy. I’d love your feedback, ideas, and comments. App Store: https://ift.tt/13R7UEY Website: https://activiews.xyz Thanks! http://activiews.xyz/ July 11, 2025 at 12:28AM
Show HN: RecomPal – A no-code AI chatbot to increase Shopify sales Hi HN! We’ve built RecomPal, a no-code AI chatbot designed specifically for Shopify stores. It helps merchants increase conversion rates and average order value by assisting shoppers in real-time—just like a human sales rep. Key features: Plug & play installation (2 minutes) Understands customer intent and recommends products No scripting or flow-building required Privacy-first: no data sharing with third parties We’ve seen up to 30% sales increase in early tests with small-to-medium stores. We’d love your feedback, feature suggestions, or ideas on how we can improve. Try it out: https://recompal.com Thanks! – Team RecomPal https://recompal.com July 9, 2025 at 10:41PM
Show HN: Stravu – Editable, multi-player AI notebooks with text, tables, diagram Hi HN! I'm Karl one of the co-founders of Stravu. ( https://stravu.com ) Using AI for work 24x7, we realized that four things would make AI more useful for us and a lot of other power users and teams: Editable output: AI gives output that is half right and our only option was to either keep chatting laboriously or copy it to a Google Doc. We made Stravu so you can edit what the AI says in chat or in an attached notebook. Everything editable. Approve AI changes: When AI makes a change to some text, you can't tell what changes. We put Red/Green diffs that you can approve into Stravu. Unify text, tables, diagrams: We were jumping between tools to work with AI on text, tables, diagrams, etc.. Just because Microsoft did that 30 years ago, doesn't mean it makes sense now. We made Stravu so you can work with AI across text, tables, diagrams, (and 2x2s, formulas, more soon) and have them inform each other. Actual multi-player team collab with AI: We couldn't collaborate as a team in AI (even with the ChatGPT Teams plan). We wanted to be able to chat with AI together as a team or see the changes AI was making in the canvas/notebook together and edit together. So we made Stravu support multi-player collaboration in every aspect... chats, notebooks, text, tables, diagrams..etc. Some of the use cases of our current Beta customers include: scrum teams doing feature/customer/competitive research and feature definition, account teams building vertical/geo/account plans, consultants and investment teams working on market/company analysis. We are currently in beta and actively iterating based on user feedback. Please try it out at: https://stravu.com We highly value your feedback! July 9, 2025 at 07:47PM
Show HN: OpenAPI mocks that don't suck – realistic test data, quick setup Hi HN! I'm Ankit, the founder of Beeceptor, a request mocking and intercepting tool. This time, I built something new to address a pain I’ve personally felt (and heard from dozens of QA, frontend, and platform teams): making OpenAPI specs actually useful during development. API contracts just sit around as docs often. What if you could 'activate' them, instantly have a realistic, hosted mock server-with contract validation, smart test data, and early usage? So I made _Mockaroo for OpenAPI_, but with brains. It spins up a hosted mock server from your spec in one click. It: - Generates sensible context-aware, test data using FakerJS (e.g., age returns realistic numbers, not 10000) - Validates incoming requests against your contract definition and returns detailed, actionable error messages. - Supports JSON, binary, CRUD style API responses. - Gives a HOSTed API server URL, that's ready in a few seconds. - Helps frontend teams start testing before the backend is ready - Gives QAs a place to play, verify and run performance tests. - Enables the best developer experience, requiring no account setup and a working mock API server for experiments, and API cost savings. No local setup, no writing custom mock rules, no fuss. Just activate your OpenAPI spec, and your API starts “working” in seconds. Besides, Beeceptor shows live request logs, where responses can be overridden. Quick demo: https://www.youtube.com/watch?v=2vUD_B3aw5I --- Would love your thoughts, feedback, or use cases I haven’t thought of yet. Happy to share more technical details if there's interest. Thanks for reading! https://ift.tt/Q06sbX2 July 9, 2025 at 01:00AM
Show HN: A rain Pomodoro with brown noise, ASMR, and Middle Eastern music I built this because most Pomodoro timers felt too sterile. I wanted something that actually pulls you in with rain, brown noise, soft ASMR, and a few Middle Eastern tracks. Added animated backgrounds so it’s not just a blank screen. Runs fully in your browser. No accounts, no tracking, just open it and focus. If you give it a try or have ideas to make it better, I’d love to hear. https://ift.tt/8TJ4lna July 8, 2025 at 11:12PM
Show HN: OffChess – 100k+ Offline, Ad-Free Chess Puzzles App Hi HN! I'm the developer of rdx, a mildly popular ad-free, privacy and user friendly Reddit client. This time, I made something for a very specific use case: solving chess puzzles with no internet. Why? Well, my Wi-Fi is terrible in the bathroom—and that's where I do some of my best thinking. I tried printing out “mate in X” puzzles to solve offline, but they weren’t fun without interaction. So I built OffChess. OffChess is an iPhone/Android app that contains over 100,000 chess puzzles, fully offline and completely ad-free. You can solve puzzles by category (Mate in 1/2/3/4/5, tactics like pins/forks/skewers, or openings like Sicilian/French, etc). You gain or lose points based on how you perform, so there's a light rating system to keep things engaging. No accounts, no tracking, no monthly subscriptions, no internet required. Just pure, old-school tactical chess training, wherever you are. You can check out the iPhone/iPad app at https://ift.tt/mK5yFXR... or the Android app at https://ift.tt/vGEhd2c Would love feedback, bug reports, or suggestions. Thanks! https://offchess.com July 8, 2025 at 02:27PM
Show HN: HireIndex – A Searchable Directory for Who Wants to Be Hired on HN Hi HN, I built hireindex.xyz – a searchable website that aggregates and indexes candidates from the "Who Wants to Be Hired" threads on Hacker News. Coming soon: Analytics to highlight trends in skills, technologies, and candidate preferences across HN posts. With HireIndex, you can: Search by tech stack. Quickly scan bios, salary expectations, and contact links. Filter by remote/on-site preferences and employment type. I made this because I was tired of scrolling through long threads and wanted a better way to find interesting people. Would love your feedback – especially around UX and anything that would make it more useful to hiring managers, founders, or even job seekers themselves. https://hireindex.xyz https://hireindex.xyz July 8, 2025 at 09:41AM
Show HN: Life_link, an app to send emergency alerts from anywhere 2-factor authentication terrorizes me. Many years ago, I was at a friend's house and decided to quickly pop downstairs to buy something. When trying to re-enter the building, the main gate had locked. And since my phone, keys, and any spare cash were all left upstairs, I was stuck outside. Thankfully, Google, at the time, allowed users to send SMSes (SMSs?) through the Hangouts app, which I did after logging into my account from an internet cafe. (Back then people weren’t constantly connected to Facebook, etc.) Since that day I feared being locked out of my accounts by 2FA, simply because my phone (and my access codes) weren't with me. And while it's true that today people are always connected to messaging apps, I wouldn't be able to reach any of them without me being able to log into mine. That’s why I built life_link, an app that allows my loved ones to reach me, wherever they are, without having to care about 2FA or even needing to log in. To do so, they simply need to reach the app on a secret (short) URL. I've since expanded life_link by creating a "generator" app that can produce a pre-configured, single-file application, ready to be deployed on any static site hosting service and decided to open-source it for other people who might find it useful. You can find the life_link project and learn more about it here: https://ift.tt/xZLlr32 July 8, 2025 at 02:29AM
Show HN: I wrote a "web OS" based on the Apple Lisa's UI, with 1-bit graphics https://ift.tt/opTwLcF This is a web OS I wrote in vanilla JS that looks like the Apple Lisa Office System (1983-85), with other contemporaneous influences and additional improvements and features. It's currently in alpha and isn't remotely bug free. I had been holding off on posting this here until it was somewhat presentable and useful. Please note; the Lisa conforms more literally to the desktop metaphor than most modern GUIs - some of the important differences are mentioned in the readme. This is a complete recreation of the UI in JS; it all renders to a single canvas element. It's not a CSS theme, and not an emulator ported to JS. None of the code is written by Apple. I'll be happy to elaborate more in the comments, but the short version is the entire UI is defined outside the DOM using JS objects. Thus, every interface element - menus, windows, controls, and even typefaces - was recreated from scratch. There are no font files - I wrote my own typesetting system, which supports combining multiple text styles and generates new glyph variants on the fly. Many of the technical decisions I made were motivated by a desire to have this look the same in every browser. That's harder to do with the DOM and CSS, and why I moved as much logic as I could to JS. Also, the only part of the project outside of vanilla JS and standard web APIs is the Gulp toolkit, which I'm using as a minification/build tool. No vibe coding was used to make this! This is based on a UI from the 80s, and won't work well on your phone. If you insist on running it that way, turn on trackpad mode in the touchscreen settings panel of the preferences app. For best results, install it as a PWA (add it to your home screen). Also there are some odd Android bugs; the native touchscreen keyboard is currently broken, and there's an issue with the cursor when dragging windows. I realize there's not a whole lot to do within LisaGUI right now; I've got a big list of additional features and apps I'll be adding in the future. I've been working on this project for a while, and I'm eager to hear people's feedback and answer questions about it. https://ift.tt/d7QhFYy July 7, 2025 at 12:02AM
Show HN: Simple wrapper for Chrome's built-in local LLM (Gemini Nano) Chrome now includes a native on-device LLM (Gemini Nano) starting in version 138. I've been building with it since it was in origin trials, it's powerful but the official Prompt API is still a bit awkward: - Enforces sessions even for basic usage - Requires user-triggered downloads - Lacks type safety or structured error handling So I open-sourced a small TypeScript wrapper I originally built for other projects to smooth over the rough edges: github: https://ift.tt/LBtgdRe npm: https://ift.tt/4Xj3Cut - Stateless prompt() method inspired by Anthropic's SDK - Built-in error handling and Result-based .Safe.* variants with neverthrow - Token usage checks - Simple initialization that provides a helper to trigger downloads (must be triggered by user action) It’s intentionally minimal for hacking and prototyping. If you need fine-grained control (e.g. streaming, memory control), use the native API directly: https://ift.tt/IiKlsWQ Would love to hear what people build with it or any feedback! https://ift.tt/LBtgdRe July 6, 2025 at 11:24PM
Show HN: From Photos to Positions: Prototyping VLM-Based Indoor Maps Just a fun hack I did while bored over the weekend. My wife was busy shopping, it got me thinking that can VLMs solve the indoor location problem in a mall? Can I just show a VLM a map and an image and have it doa good enough job locating me? I hacked this P.O.C and it seems to work. https://ift.tt/J1cpCRm July 6, 2025 at 04:49AM
Show HN: Certimon – Free Telegram bot to monitor SSL certificate expiry Hey folks, I'm a developer who got tired of having to rely on SaaS providers for such a simple task as monitoring an SSL certificate expiry. I had to create calendar entries just for the sake of making sure we don't forget it because of course warning emails got lost or the person responsible was on holiday.. So instead of "building yet another platform" I decide to take the un-sexy and simple route and put together a Telegram bot, called it Certimon. Thanks to Claude and Cursor I also put together a landing page for it https://certimon.com/ but I don't think you'll get any value out of it... If you want to give it a try, just add @CertimonBot to your Telegram or visit https://ift.tt/79KnMdJ to add directly. You can check a domain like this: /check google.com And get reminders like: /remind google.com 30 It's 100% free, supports as many domains as you need and you won't need to sign up to anything or download anything (assuming you already have Telegram). Hope someone other than me will find this useful, would love to hear your thoughts, comments and questions! Edit: Forgot to mention, but the operational cost of this solution is around £3 per month on AWS. -Tamas https://certimon.com/ July 6, 2025 at 03:29AM
Show HN: Distapp. Manage and distribute Android, iOS and Desktop app Hi HN, I built DistApp, a tool for managing and distributing Android, iOS, and Desktop app builds. I created it after App Center Distribution was discontinued, I wanted a way to easily share builds with the QA team and users with different groups. DistApp lets you manage multiple apps across different organizations and groups. It also supports self-hosting on your own infrastructure and is open-source https://ift.tt/xQLrPuY You can try the free version by signing in with your Google account. I chose not to support email/password accounts to reduce abuse on the free version. But I’m open to suggestions if you think there’s a better way :) Thank you https://ift.tt/pFWDZEn July 5, 2025 at 11:54PM
Show HN: SteadyText: Deterministic LLMs: Same input → same output, every time Hey HN! After spending way too many nights debugging flaky AI tests, I built SteadyText. It's a simple python library for deterministic llm generations and embeddings. We use it in production for: - Testing our AI features (zero flakes in 3 months) - CLI tools that need consistent outputs - Reproducible documentation examples It's not for creative tasks - this is specifically for when you need AI to be boring and predictable. Think of it as the opposite of ChatGPT. The coolest part? It includes a Postgres extension. You can now do: SELECT steadytext_generate('explain this query: ...'); And it will always return the same explanation. :) How it works: 1. Greedy decoding- Always pick the highest probability token (no randomness) 2. 8-bit quantization- Same numerics across all platforms It's easy to get started: uv tool install steadytext echo Hello | st https://ift.tt/NGt0p7u July 4, 2025 at 12:27AM
Show HN: A browser extension to control Google's Random Number Generator TLDR: Built a simple extension to see how easily client-side logic can be manipulated. It lets you pre-select the winner of a basic online raffle as a sort of "magic trick." First time posting a project on HN. First public github repo also. Keen on feedback. I was exploring how browser extensions can interact with and modify a page's state in real-time, specifically what kind of access was needed to intercept a process without breaking the UI. The test tool was a bit of fun on its own, so I polished it up and decided to post it here. https://ift.tt/bKDs4EB July 3, 2025 at 11:20PM
Show HN: Bookmark and organise your mobile links with ease with this free app Do you have lists scattered all over your phone? Are you tired of saving recipes, books or restaurants in Notes, screenshots or Whatsapp groups? Listee is the bookmark tool to privately save and structure the things you love. Never lose track anymore of the places you loved, movies you wish to see or shoes you want to buy. Save any content in seconds using the share function on your phone or the search engine within Listee. Connect with your friends to share your favourites or create lists together. Listee is the new way to save, share and explore with the ones you love and trust. Also for your wish lists! https://ift.tt/dIQXsKa July 3, 2025 at 02:13PM
Show HN: I made a Chrome extension to export web element to code Recently I'm working on CopyUI which is an extension to copy UI element from websites and export html(or jsx) and css(or tailwind). I'm building this tool in order to create better landing pages because I'm really not good at layout and colors. So I hope to learn from others' design and innovate later, not to simply replicate. https://copyui.online July 3, 2025 at 07:32AM
Show HN: Issue Duration Labeler – a GitHub Action that labels issue by age I’ve built *Issue Duration Labeler*, a GitHub Action that automatically adds *color-coded duration labels* to every issue in a repo: Default label thresholds: Green – ≤ 7 days (configurable) Orange – ≤ 30 days Red – > 30 days For open issues we compute “age” (creation → now). You can adjust the day thresholds and label colors in the workflow file, and choose whether labels update daily or only when they cross the next threshold. *Why?* I often lost track of how long tickets had been lingering, especially in older projects. A quick glance at the issue list or github project now tells us what’s fresh, what’s getting stale, and what’s officially ancient. It’s also handy for post-mortems: sort by red labels to see which bugs took the longest to close. *Link* https://ift.tt/vnCLS16... https://ift.tt/l5hGZO6 July 3, 2025 at 03:15AM
Show HN: Runik – Turn fan wikis into e-reader dictionaries Hey HN! As a reader of epic fantasy and sci-fi, I often find myself reaching for my phone to look up an obscure side character — or the difference between “Genebackis” and “Genebaris”. So I built runik to bring in-world definitions directly onto my e-reader and stay immersed in the story. Runik parses the contents of fan wikis into Kobo and Kindle compatible dictionaries. It uses the device’s built-in word lookup feature, so there’s no jailbreaking required and definitions can be used offline. It’s still in early development and is built with Go (Wails) + Svelte + Dictutil — feedback is appreciated! Note: Kindle support requires kindlegen, which comes bundled with the Kindle Previewer app (details in the README). https://runik.app/ https://ift.tt/pdGwxmY https://ift.tt/pdGwxmY July 1, 2025 at 11:25PM
Show HN: Arch-Router – 1.5B model for LLM routing by preferences, not benchmarks Hi HN — we're the team behind Arch ( https://ift.tt/d8WOGUm ), an open-source proxy for LLMs written in Rust. Today we're releasing Arch-Router ( https://ift.tt/SPGxdCa ), a 1.5B router model for preference-based routing, now integrated into the proxy. As teams integrate multiple LLMs - each with different strengths, styles, or cost/latency profiles — routing the right prompt to the right model becomes a critical part of the application design. But it's still an open problem. Most routing systems fall into two camps: - Embedding-based routers use intent classifiers — label a prompt as “support,” “SQL,” or “math,” then route to a matching model. This works for simple tasks but breaks down in real conversations. Users shift topics mid-conversation, task boundaries blur, and product changes require retraining classifiers. - Performance-based routers pick models based on benchmarks like MMLU or MT-Bench, or based on latency or cost curves. But benchmarks often miss what matters in production: domain-specific quality or subjective preferences like “Will legal accept this clause?” Arch-Router takes a different approach: route by preferences written in plain language. You write rules like “contract clauses → GPT-4o” or “quick travel tips → Gemini Flash.” The router maps the prompt (and conversation context) to those rules using a lightweight 1.5B autoregressive model. No retraining, no fragile if/else chains. We built this with input from teams at Twilio and Atlassian. It handles intent drift, supports multi-turn conversations, and lets you swap in or out models with a one-line change to the routing policy. Full details are in our paper ( https://ift.tt/cLgxMbs ), but here's a snapshot: Specs: - 1.5B params — runs on a single GPU (or CPU for testing) - No retraining needed — point it at any mix of LLMs - Cost and latency aware — route heavy tasks to expensive models, light tasks to faster/cheaper ones - Outperforms larger closed models on our conversational routing benchmarks (details in the paper) Links: - Arch Proxy (open source): https://ift.tt/d8WOGUm - Model + code: https://ift.tt/SPGxdCa - Paper: https://ift.tt/cLgxMbs July 1, 2025 at 10:43PM
Show HN: Lifp – A Lisp Built on Bun A silly summer break project where I played around with Bun, which I hadn't had the chance to yet. It's not super lisp-y, there's no car, no cdr, nor macros. It's a Lisp-Inspired Functional Programming language that ideally doesn't require a paradigm shift when you come from C-like languages. https://ift.tt/qKQPHp0 July 1, 2025 at 08:04PM
Show HN: Audiopipe – Pipeline for audio diarization, denoising and transcription Audiopipe is a one-liner for denoising, diarization and transcription with demucs + pyannote + insanely-fast-whisper. Made it to transcribe podcasts and Dungeons And Dragons sessions, figured it could be useful. It also has a wasm UI to load transcriptions and audio. Feel free to contribute! Feedback appreciated. https://ift.tt/ThewoKm July 1, 2025 at 01:02AM
Show HN: We're two coffee nerds who built an AI app to track beans and recipes It’s available on iOS now: https://ift.tt/5RQJ0Kj We got into specialty coffee during COVID and, like many others, fell deep down the rabbit hole. Along the way, we ran into the same frustrations: - A drawer full of empty coffee bags. - No simple way to track grind size, rest dates, notes—by bean. - My coffee history scattered across photos, screenshots, notebooks, and half-memories. - The unique traits, people, and stories behind each coffee disappearing from the internet once it sold out (since coffee is an agricultural good) - In our opinion, no coffee tool really captures the flavor, emotion, and aesthetic of great coffee—from a design perspective. So we built BeanBook—a coffee notebook log beans, extract recipes, and organize your coffee life in one place with just a snap, powered by AI Here’s what it does: - Snap a bag → Auto-detects roaster, origin, process, roast date, notes, producer, farm, and more - Paste a YouTube link or photo → Extracts a structured recipe automatically - Log grind size, roast timeline, ratings & notes → All saved in a clean, elegant UI - See your coffee year in review → Track habits, trends, and favorites - Ask BeanBook AI → From brew temps to bean facts, get instant answers My co-founder and I built everything ourselves—branding, code, and UX design. If you’re into coffee (or trying to get more into it), we’d love your feedback. - Rokey & Eric https://beanbook.app July 1, 2025 at 12:08AM
Show HN: QuizKnit, an open source quiz creator QuizKnit is an open source quiz creator to create quizzes by pasting in text or selecting from a large range of historical categories. Started as a very simple app for me to play around with OpenAI’s API last year then morphed into a portfolio project during my job search earlier this year. Now happily employed so decided to open source it. Right now, a user can create a quiz, take a quiz, save it and share the quiz with other people using a URL. You can try out the full working application at https://quizknit.com Github Links: Frontend: https://ift.tt/ZaFwSCr , Backend: https://ift.tt/I1ZzPEp https://quizknit.com June 30, 2025 at 10:46PM
Show HN: Ciara – Securely deploy any application on any server Hey HN! Coolify and Kamal were "nice" (Kamal docs are pretty bad, actually), but I still had to configure firewalls, unattended-upgrades, and Fail2ban every single time. Ciara does all of this from a single configuration file. Features: Integrated Firewall Automatic System Updates Zero-Config OS Ready Zero-Downtime Deployments Automatic HTTPS support Multiple Servers Deployments Would love your feedback and happy to answer any questions! https://ift.tt/pjYCfdH June 30, 2025 at 01:30AM
Show HN: Anti-Cluely – Detect virtual devices and cheating tools on exam systems Anti-Cluely is a lightweight tool designed to detect common virtual environments, device emulators, and system manipulation tools often used to bypass or cheat in online exams. https://ift.tt/daS8zDN June 29, 2025 at 01:11AM
Show HN: Query your Rust codebase and generate types for anything Hello HN! As a long-time professional Rust developer. I've always been frustrated by the difficulty and "hackiness" of producing bindings to other languages, whether a frontend, an FFI library, etc. Not just in Rust but in any language. After many years of trying existing solutions and trying to make my own, I've finally developed a solution I'm very happy with. RTK (aka Rust Type Kit) allows you to write Lua scripts that perform queries on your code, such as method calls to Axum's `.route`, function definitions, and more, and then receive rich type information including all argument types, function paths, proc macro attributes, and more. Your Lua script can then read this information and emit an output file in any language of your choosing. Or, you can emit compiler errors and use it as a linter of sorts. You can even directly re-emit Rust code itself and use this as a richer proc macro solution! The code example is a bit verbose, so I encourage you to take a look at the repo's README. I look forward to hearing your thoughts, or any usecases you may come up with! https://ift.tt/JaLb32T June 29, 2025 at 12:25AM
Show HN: I built Hispi, an app to design custom jewellery Designing is free and gives you a breakdown of material choices and costs, then you can get a quote if you want to buy the ring. I'm working with around 50 jewellers in London and the UK's Jewellery Quarter but I think just designing a ring is pretty nice. https://hispi.app June 24, 2025 at 07:08PM
Show HN: IssuePay – Get paid for open-source contributions Hi HN! I’m Mario, and I’m about to launch IssuePay. Problem: Open-source contributors don’t get direct financial recognition for their work. Solution: IssuePay lets maintainers post bounties on GitHub/GitLab issues. Contributors pick tasks, merge code, and get paid automatically. You can then withdraw your earnings directly to your Bank Account. Try it out: https://issuepay.app Questions: Would love feedback on our UX, payout reliability, or any scaling tips. Note: Open to partnerships with OSS communities! Thank you, guys ! <3 https://issuepay.app June 28, 2025 at 12:01AM
Show HN: I made a tool to convert recipes to an easy reading graphical notation Hi HN, I have been working on a project which can automatically convert a recipe to the tree/graph/tabular style notation that pops up on HN every now and then. It's still early days but I am keen to hear about improvements I can make to the recipe presentation. For reference, the earliest example I have seen of this style of notation is from 'Cooking For Engineers' ( https://ift.tt/Pok2a4y ) but I have seen other, similar notations from time to time. This is my take on the notation, as well as the ability to convert recipes automatically. In the future, I would like to see it become my own personal kitchen/recipe management system but I would first like to improve the recipe display options. It uses ChatGPT to format the recipe into a JSON tree structure, which is then rendered with CSS grid. So far it works well enough but there is definitely some recipes that are impossible to render, and of course ChatGPT will occasionally spit out less-than-useful trees. I have been using AppWrite for the backend. Curious to hear everyone's thoughts and what people might like to see! https://gobsmacked.io June 27, 2025 at 11:10PM
Show HN: Pocket2Linkding – Migrate from Mozilla Pocket to Linkding With the Mozilla Pocket shutdown coming up in about two weeks, I thought I'd share this quick tool to migrate to linkding in case it's helpful to others. After reviewing self-hosted options to Pocket, I decided linkding has the best combination of features. (The creator/author of linkding has done a great job -- however, I plan to eventually create a new tool that is based on linkding but adds some new features that the author has indicated he doesn't want to include [I’m currently using a fork, but I want to expand on it further].) HN thread about shutdown announcement: https://ift.tt/qVSDYfv Mozilla announcement: https://ift.tt/OwY8CGq linkding: https://linkding.link/ Note that Pocket is shutting down July 8, 2025, but the export service will remain available until October 8, 2025. [edit] fix typo in title & formatting https://ift.tt/Z0tVczU June 26, 2025 at 10:33PM
Show HN: Chat with ChatGPT, Claude, Grok, Gemini and Llama on One UI Chat with multiple AI models once and compare the results to pick the best one. This should help you with your research as different AI model can give you different answers and some might be better than others. https://instaask.ai June 26, 2025 at 11:12PM
Show HN: I rawdog a MCP server from scratch in Zig. No SDK Some time ago I wanted to write a MCP server in Zig but found out there's no real JSON-RPC support in Zig, which MCP needs for communication. I ended up developing a JSON-RPC 2.0 library in Zig and more [1], which had its challenges. So I finally was able to put together a MCP server in Zig. It's built from scratch implementing the protocol messages from the MCP JSON schema. It's actually quite magical to have the LLM calling my MCP server [2]. The work is not too bad. Most of the hard work has already been done in the JSON-RPC library. [1] https://ift.tt/B2GwSWe [2] https://ift.tt/znmUr9I... https://ift.tt/JwCPW6m June 25, 2025 at 11:44PM
Show HN: Logcat.ai:AI-powered observability for Operating Systems(Android+Linux) Hello HN! I'm an Android OS engineer. I've worked with AOSP and Linux kernels all my career and always wondered about lack of sophisticated tools to debug and analyze system-level logs. Always had to resort to manually skimming through large log files to find something I needed to. With the rise of LLMs and the AI-age, I felt it was a great opportunity to build something for OS engineers, which is what led to logcat.ai! We are building the industry-first observability platform for system level intelligence. Think "Datadog for operating systems" instead of applications. Currently, we support Android and Linux - more platforms on the way. With Android we offer: 1. logcat analysis: Ability to analyze logcat logs for root cause analysis of system issues with natural language search. Unlike, Firebase which is an app-level observability, logcat.ai provides intelligence at OS level spanning bootloader, kernel and framework layer. 2. bugreport analysis: As you know a bugreport is a super-verbose snapshot of an Android OS collected at a point of time. Analyzing these logs takes hours and sometimes even days. We are working to bring this down to minutes! Analysis of memory, cpu, process stats to infer memory pressure levels, system stress, and nail down the processes responsible for it, identify performance bottlenecks and memory leaks across the system. For Linux we offer: dmesg (kernel log) analysis to help identify issues at Linux kernel level. We plan to add support for different Linux distros with their own logging pretty soon. Our goal is to build a single-pane-of-glass observability experience for operating systems worldwide, something that's never been done before. Our website may not reflect all the features a.t.m but we have a lot of things cooking! Ask us anything. We are providing free beta access for a period of time. We'd love your feedback and comments on what you think about logcat.ai! https://logcat.ai June 24, 2025 at 10:53PM
Show HN: TNX API – Natural Language Interactions with Your Database Hey HN! I built TNX API to make working with databases as simple as asking a question in plain English. What it does: - You write a natural language prompt (e.g., "List products with price > 20 USD") - Our system turns it into SQL and runs it - You get actual results, optionally visualized - Your data stays private – nothing is stored, the AI doesn‘t see it, and the API forgets immediately after replying Why I made this: Writing SQL for routine questions is https://ift.tt/kHTUjPl still a blocker for many teams. I wanted a privacy-first, plug-and-play API that just works with natural language. TNX doesn’t just translate — it executes the queries and returns actual answers (not just SQL). Examples: - You ask: “Total sales by product category this year?” → TNX replies: [furniture: $43,000, electronics: $12,000] + “Want a chart for this?” - You ask: “Which customers didn’t order in the last 90 days?” → TNX replies with names or IDs and offers follow-up actions Notes: - Built on modern AI models (small + fast) - No need to send full database dumps – just metadata/config + real-time access - Easy API integration - (Bonus: If you should be interested, I‘d handle setup + customization for you) Try it out: https://ift.tt/kHTUjPl (user name: „hi@tnxapi.com“, password „1“ (so it's harder to forget)) (example promts: - „Please give me the name, ShortDescription and price of product with idpk = 20.“ or - „Please list me all product prices from idpk 10 to 20.“ and then - „Please list me all product prices from idpk 10 to 20.“ (I copied some of my databases for this test, I am sorry for the data being in German xd)) Cheers, Lasse Tramann (Feel free to reach out to hi@tnxapi.com : ) ) https://ift.tt/kHTUjPl June 24, 2025 at 12:48AM
Show HN: Pickaxe – a TypeScript library for building AI agents Hey HN, Gabe and Alexander here from Hatchet. Today we're releasing Pickaxe, a Typescript library to build AI agents which are scalable and fault-tolerant. Here's a demo: https://ift.tt/aix7qYd... Pickaxe provides a simple set of primitives for building agents which can automatically checkpoint their state and suspend or resume processing (also known as durable execution) while waiting for external events (like a human in the loop). The library is based on common patterns we've seen when helping Hatchet users run millions of agent executions per day. Unlike other tools, Pickaxe is not a framework. It does not have any opinions or abstractions for implementing agent memory, prompting, context, or calling LLMs directly. Its only focus is making AI agents more observable and reliable. As agents start to scale, there are generally three big problems that emerge: 1. Agents are long-running compared to other parts of your application. Extremely long-running processes are tricky because deploying new infra or hitting request timeouts on serverless runtimes will interrupt their execution. 2. They are stateful: they generally store internal state which governs the next step in the execution path 3. They require access to lots of fresh data, which can either be queried during agent execution or needs to be continuously refreshed from a data source. (These problems are more specific to agents which execute remotely -- locally running agents generally don't have these problems) Pickaxe is designed to solve these issues by providing a simple API which wraps durable execution infrastructure for agents. Durable execution is a way of automatically checkpointing the state of a process, so that if the process fails, it can automatically be replayed from the checkpoint, rather than starting over from the beginning. This model is also particularly useful when your agent needs to wait for an external event or human review in order to continue execution. To support this pattern, Pickaxe uses a Hatchet feature called `waitFor` which durably registers a listener for an event, which means that even if the agent isn't actively listening for the event, it is guaranteed to be processed by Hatchet and stored in the execution history and resume processing. This infrastructure is powered by what is essentially a linear event log, which stores the entire execution history of an agent in a Postgres database managed by Hatchet. Full docs are here: https://ift.tt/LvZJ5NC We'd greatly appreciate any feedback you have and hope you get the chance to try out Pickaxe. https://ift.tt/lo1D02a June 20, 2025 at 09:37PM
Show HN: A Tool to Summarize Kenya's Parliament with Rust, Whisper, and LLMs Bunge Bits summarizes long parliamentary sessions from the Kenyan National Assembly and Senate. Built with Rust, Whisper v3, and GPT-4o. Sessions are typically 3–7 hours long, mixing English and Swahili. This tool transcribes, chunks, and summarizes them to make political content more accessible and searchable for the public. https://ift.tt/6WsFVSc https://ift.tt/XYoLFWj June 22, 2025 at 11:03PM
Show HN: Remotely Good – AI-powered job platform for remote, mission-driven work Hi HN! I’m Theresa, founder of Remotely Good, a one-stop platform to help people find remote and hybrid jobs with mission-driven orgs—nonprofits, campaigns, social enterprises, and public agencies. Remotely Good offers: -Curated job listings by salary, cause area, and location -AI-powered career tools (resume enhancer, job matching, cover letter gen, interview prep) -A coaching marketplace (coming soon!) for affordable 1:1 guidance -Volunteer roles, activism opps, org culture insights, and more I’ve built MVPs of several tools, and I’m now validating interest in upcoming features—including an AI voice career coach and more. I’m looking for at least 50 early users to sign up for premium access and beta tools at remotelygood.substack.com Feedback is gold—please check it out and fill out this short survey: https://ift.tt/KHFx0h4 Try it out: https://ift.tt/I8mdcBz Feedback survey: https://ift.tt/KHFx0h4 Twitter/IG: @remotely_good Would love your thoughts on: Are these tools actually helpful to social impact jobseekers? Any features I’m missing? How could I better reach first-gen and mission-driven users? Thanks for reading! https://ift.tt/I8mdcBz June 22, 2025 at 10:56PM
Show HN: Good old emails and LLMs for automating job tracking So I spent the last few days building Jobstack. The logic is quite simple. You apply to jobs and you get emails, you trade emails back and forth from interviews, questions and others until the role is either accepted or you are rejected. Also easy to apply to hundreds of roles and not being to know where you stand easily. With Josbtack, you sign up, get a unique email and forward emails to the url. And it uses LLMs to extract company details , tries to find information online about them and presents that to you. Every email you forward becomes part of your timeline with the company. It also tracks rejection, offers from the emails too and gives you a nice stats dashboard amongst others. Using Gemini 2.5 pro right now. No data stored not in any way. After extraction, it’s discarded. Even “AI chats with the company” aren’t stored https://jobstack.me June 22, 2025 at 03:07AM
Show HN: Swift UI app for extracting beer information by just taking photos I would like to share Swift UI app for extracting beer information by just taking photos. It is based on Gemini API and you can easily use this as reference to create an AI supported iOS app. https://ift.tt/A3EL7X2 June 21, 2025 at 03:49PM
Show HN: Inspect and extract files from MSI installers directly in your browser Hey everyone! I'm excited to share a small web app I built that allows you to view and extract the contents of Windows MSI installers directly in your browser. It's essentially a web-based "lessmsi" powered by Pyodide. You can try it out at: https://ift.tt/I6WDVuA My motivation for building this was from part of my day job -- I often get Windows MSI installers and need to extract files while preserving the relative directory structure and filenames, as they would appear after a full installation. The existing tools I found were good but limited in which platforms they support: lessmsi works great on Windows, while msitools works for Linux/macOS. Neither is a truly cross-platform solution that works on any major OS. So we developed pymsi (a pure Python library, available on GitHub at https://ift.tt/1PBpQOs ) to handle reading and extracting MSI files from Python. Then I realized that since pymsi has no native dependencies, it could potentially run in a web browser using Pyodide. After a bit of "vibe coding" and fixing some "hallucinated" functions/classes that don't exist in pymsi, the result was this client-side web app. If you need an MSI file to experiment with, older versions of PowerToys included the installer in .msi form, such as this one: https://ift.tt/np1REFD.... Note that the underlying pymsi library hasn't been extensively tested against a bunch of MSI installers yet, so there might still be lingering bugs. If you come across any issues, please don't hesitate to report them in on the GitHub repository ( https://ift.tt/egO1mNQ ). I'd love to hear your feedback and answer any questions! https://ift.tt/I6WDVuA June 21, 2025 at 01:34AM
Show HN: Vpuna AI Search – A semantic search platform Dear HN Community, I am a long time fan and first-time contributor. I just launched a developer focused semantic search platform and wanted to share it with the community. The idea is simple: upload structured or unstructured documents, select the fields you want to index and tag as metadata, and instantly get a clean search API you can use in your own app. Here is what it currently supports: - Manage your own tenants and projects - Upload .json and .txt files (support for .pdf, .docx, .xlsx, .yml, etc. coming soon) - Expose 3 APIs: search, upload document (embeddings), and delete document - Manage your own API keys - Uses CPU based sentence-transformers/all-MiniLM-L6-v2 for embeddings ( support for other local and online models are coming soon) LLM summarization and Model Context Protocol (MCP) support are on the roadmap Why I built it: In my consulting work, I kept seeing client wanting to move beyond basic keyword search and integrate semantic search with optional summarization. Most existing tools are either too expensive, too restrictive, or require custom layers (like custom Python servers for pre processing queries and embeddings). I wanted something API first, developer friendly, and easy to self host or use out of the box. This is the first release, and I would love your feedback. Would you use this? What is missing for your use case? Here is the README with all the links https://ift.tt/7G9jS8k Thank you for your time. https://ift.tt/Gu7UkpK June 20, 2025 at 11:24PM
Show HN: Gaussian Random Walker Simulation in JavaScript Was going through Nature of Code and came across the idea of Gaussian Random number generator, so build a simulation that generated random walkers who walk based on this and also the walkers are generated based on random numbers from a gaussian distribution. Added additional features and toggles that make it possible to create art (like setting persistent to true), colors, exporting as gif and image. https://ift.tt/whOljQi June 20, 2025 at 01:15AM
Show HN: EnrichMCP – A Python ORM for Agents I've been working with the Featureform team on their new open-source project, [EnrichMCP][1], a Python ORM framework that helps AI agents understand and interact with your data in a structured, semantic way. EnrichMCP is built on top of [MCP][2] and acts like an ORM, but for agents instead of humans. You define your data model using SQLAlchemy, APIs, or custom logic, and EnrichMCP turns it into a type-safe, introspectable interface that agents can discover, traverse, and invoke. It auto-generates tools from your models, validates all I/O with Pydantic, handles relationships, and supports schema discovery. Agents can go from user → orders → product naturally, just like a developer navigating an ORM. We use this internally to let agents query production systems, call APIs, apply business logic, and even integrate ML models. It works out of the box with SQLAlchemy and is easy to extend to any data source. If you're building agentic systems or anything AI-native, I'd love your feedback. Code and docs are here: https://ift.tt/XMymSAV . Happy to answer any questions. [1]: https://ift.tt/XMymSAV [2]: https://ift.tt/0jERZ1O https://ift.tt/XMymSAV June 19, 2025 at 11:02PM
Show HN: Simstack, SSH escape room for developers I built a realistic prod system under heavy pressure where you SSH in and solve real problems. You get your own server (real box) and have to fix various things up under fire. a friend told me it reminded them of DOOM for SREs, another called it a "flight simulator" because you can do anything (and reboot if you crash). someone else said it was like "an escape room for engineers", in that you have to find your own way around an unfamiliar box and solve puzzles. idk, maybe it's fun. try it! how it works: you get two Hetzner servers: a traffic generator (real traffic from NYC taxi data highly compressed to 12k rps) firing tons of requests at a user server, which has some (realistically) broken stuff trying to get that data to a chart. your job is to get it working and get the whole system latency down. You can use whatever tools or techniques you like to do so. so far I've seen people messing around in databases, adding caching, rewriting services in go. There's no one right solution! why I'm building it: I was a school teacher 15 years ago and have been training developers for the last 10. most dev training (tutorials, toy projects &c) feels nothing like the "real job". I wanted to see if we can make training more realistic, challenging, fun. https://simstack.io June 19, 2025 at 10:03PM
Show HN: VerifyWise, an open-source governance platform for AI compliance Hi all, Gorkem here. I started VerifyWise [1] to make AI governance less painful. Today, we’re launching our open-source platform to help teams take control of their AI compliance process. VerifyWise helps organizations navigate AI governance by providing audit readiness, risk registers, model fairness checks, and compliance documentation. Those are all built into a single platform you can self-host. We’ve been quietly building VerifyWise for a while, and we’re now at a place where it’s ready for more teams to try. Since we started, we've: - Released our core platform on GitHub: https://ift.tt/7uoYFjh - Added workflows for EU AI Act and ISO 42001 with exportable audit trails - Built a bias & fairness module with Fairlearn integration - Developed a training registry for tracking internal AI literacy - Shipped risk register and vendors/vendor risks module - Opened our Figma for design contributions We’re aiming for full transparency and community-led development. If you're building with AI and feeling the pressure of upcoming regulations, we’d love to hear your feedback. [1] GitHub: https://ift.tt/7uoYFjh [2] Features: https://verifywise.ai [3] Documentation: https://ift.tt/koR3jQw https://verifywise.ai/ June 18, 2025 at 11:17PM
Show HN: I built a FOSS tool to run your Steam games in the Cloud I wanted to play my Steam games but my aging PC couldn’t keep up, so I built Cloudy Pad - a tool to run Steam in the Cloud (GitHub: https://ift.tt/3ZJA7nV ) It runs on AWS, Azure, GCP, Scaleway and Paperspace with various cost optimizations and safeties: - Cost alerts - Auto stop inactive instances to avoid unwanted cost - Disk snapshots and data cleanup for cost efficiency - Spot instance support Under the hood: a Linux VM and a container running Sunshine (a streaming server https://ift.tt/O1f82eZ ) with Steam. Most Windows games work just fine thanks to Proton. It streams effortlessly at 1080p 100+ FPS - I recently played Baldur’s Gate III and Clair Obscur in Ultra, ran like a breeze. Cost-wise it’s great for occasional players: ~30h or less per month typically cost less than 25$. Though admittedly for heavy gamers it may be less cost-effective due to cloud pricing. I’d love feedback from the HN community ! https://ift.tt/3ZJA7nV June 18, 2025 at 12:27AM
Show HN: TagLib-Wasm, a TypeScript-first music tagging library I was RIF'd last week so I finally had the chance to try some non-AIDE solutions (e.g. OpenAI Codex, Google Janus, Claude Code) for the purpose of building something non-trivial. I decided to create a CLI utility related to the care and feeding of my music library. Almost instantly, I ran into a problem: There are no complete music metadata libraries for TypeScript/JavaScript runtimes. TagLib ( https://taglib.org/ ) looked great, but I couldn't find any good TypeScript wrappers for it, and I wanted to avoid including OS-specific binaries in a library that might be distributed as part of standalone tools. Unfortunately, the TagLib project doesn't (yet) include a native Wasm build. (Note: I'm not a professional software engineer. My roles are historically developer relations, product management/strategy, and product marketing. I have a computer science background and have helped lead teams building extremely complex systems, so this project is somewhere between "real software engineering" and "vibe coding".) After some research and experiments, I settled on a stack of Deno, Wasm, Emscripten, and Emscripten Embind (which I had never heard of before starting this project), with Claude Code as my "pair programmer". I've built stuff with TypeScript (targeting Node.js as the runtime) before, but this was my first experience with everything else. Claude Code has surprised me several times during the development of TagLib-Wasm. Here's a response to questions I had about an earlier suggestion it had made regarding improving type safety: https://ift.tt/3ocI4CD Here's why I think TagLib-Wasm is interesting: → It's the only complete library for any-format music metadata management for TS/JS developers. → This is still somewhat aspirational until I create a suite of tests for each runtime, but I’m targeting true cross-runtime support — Deno, Node.js, Bun, Electron, Cloudflare Workers, and browsers. → I’m not aware of another library that can operate as easily with memory buffers as with files. Surely there must be one, but I suspect it’s unique in TS|JS land. After starting this, I ran into the creator of go-taglib (sentriz/go-taglib, not wtolson/go-taglib) on HN, who had also arrived at a Wasm-based solution for his Golang library. I then borrowed aspects of his developer experience to provide a "Simple API" variant for more casual use of the library. https://ift.tt/DjO42Sa June 17, 2025 at 09:58PM
Show HN: Chawan TUI web browser A terminal-based web browser in Nim.[1] Has acceptable (YMMV) CSS rendering, some JS support, and inline images (sixel/kitty). It can also use various protocols other than http(s) such as (s)ftp, gopher, gemini, ... Chawan started out as a w3m clone, and the UI still resembles it. However, the architecture has turned out quite different, with pages loaded in separate processes, and protocol/file type handling separated out into external binaries. An interesting result is that you can even register decoders for custom inline image formats, although practical use cases of this are rather minimal. There is a gallery showcasing some websites being rendered here: https://ift.tt/tXuzYsT [1]: https://nim-lang.org https://ift.tt/hBREP2c June 17, 2025 at 02:18AM
Show HN: Nexus.js - Fabric.js for 3D I was looking for a tiny library to easily transform both 2D & 3D objects with simple mouse / touch controls and a fixed camera, in the browser. Like a simple 3D editor but without requiring the user to be a Blender expert. Couldn't find anything lightweight, so I’m building one. Think Fabric.js but for 3D. Built entirely with Three.js / R3F. Borrowed some inspiration from VR/AR interaction systems for controls. Feel free to play with it and let me know what you think! https://ift.tt/Ycd3qAE June 17, 2025 at 02:03AM
Show HN: Drop – Svelte like compiler for Web components I've built a small compiler, heavily inspired by Svelte, that leans on modern web standards and proposals, namely Web Components, HTML Modules, and Signals. Although web components never really took off, I still believe they have strong potential as a foundation for building web applications without relying on a framework. GitHub: https://ift.tt/nvPol3L Blog post: https://ift.tt/MxZU0ch... I’d appreciate some feedback before committing more time to this project ! https://ift.tt/nvPol3L June 17, 2025 at 12:25AM
Show HN: Canine – A Heroku alternative built on Kubernetes Hello HN! I've been working on Canine for about a year now. It started when I was sick of paying the overhead of using stuff like Heroku, Render, Fly, etc to host some web apps that I've built. At one point I was paying over $400 a month for hosting these in the cloud. Last year I moved all my stuff to Hetzner. For a 4GB machine, the cost of various providers: Heroku = $260 Fly.io = $65 Render = $85 Hetzner = $4 (This problem gets a lot worse when you need > 4GB) The only downside of using Hetzner is that there isn’t a super straightforward way to do stuff like: - DNS management / SSL certificate management - Team management - Github integration But I figured it should be easy to quickly build something like Heroku for my Hetzner instance. Turns out it was a bit harder than expected, but after a year, I’ve made some good progress The best part of Canine, is that it also makes it trivial to host any helm chart, which is available for basically any open source project, so everything from databases (e.g. Postgres, Redis), to random stuff like torrent tracking servers, VPN’s endpoints, etc. Open source: https://ift.tt/9IplYe1 Cloud hosted version is: https://canine.sh https://ift.tt/9IplYe1 June 16, 2025 at 11:57PM
Show HN: Seastar – Build and dependency manager for C/C++ with Cargo's features Hi hackers! I'm a self-taught solo teenage dev working on Seastar, a unified build system and dependency manager for C and C++. It is capable of compiling and linking projects, managing recursive dependencies and headers, and even has a template system -- your C++ library is one `seastar new mylib --lang c++ --lib` away! Also, everything is configured in TOML, because TOML is awesome. C is one of my favorite languages, but I usually end up writing stuff in Rust because I love Cargo. Unlike C, Cargo handles the dependencies, linking, globbing, and so much more for you. So I wrote Seastar to give that function in C and C++. What's planned? A package registry like crates.io, compatibility with CMake projects, commands to migrate, and so much more. If you have more ideas, please give them! I am trying to reach 150 stars by the end of summer, and thus a star would be greatly appreciated! This project is still in development, and a star helps out a ton. https://ift.tt/W3daAJZ June 16, 2025 at 01:06AM
Show HN: Get nutrition labels for any recipe (YouTube, Website, Text, Image) Heya HN, I love watching cooking videos on YouTube, and one day an idea sparked: Can I instantly get the nutritional values for this recipe? The problem: Great recipes are everywhere, but figuring out the actual nutrition is a chore. Most of us who track calories or macros have to: * Manually get nutritional info for every ingredient. * Wrestle with spreadsheets for calculation. * Or just give up and eat that lasagna. Let’s be real: no one enjoys that, especially when you just want to cook and eat your food. So I built Recp.ai. It’s a free tool that pulls the nutrition data for you. Just give it a YouTube link, a recipe from a website, a photo of a cookbook page, or even just pasted text. It identifies the ingredients and quantities, matches them against the USDA database, and gives you a full nutrition label. It started as a script to pull ingredients from YouTube transcripts using Gemini. Then I got obsessed. Why not any website? So I added a scraper. What about any list of ingredients? Added text parsing. How about a cookbook? Now it uses Google Cloud Vision so you can just upload or snap a photo of the recipe. I wanted to build something lightweight, fast, and simple that you'd actually use. No sign-ups. No ads. Privacy-first. What I’m happy with? It works on a huge variety of sources. I fed it a photo of my old Escoffier cookbook recipe and voila - it works. What's next? I'm planning to use Google Cloud Vision to identify the dish, say "Beef Pho", it'd figure out its typical ingredients, and generate an estimated nutrition label. It would then ask users to confirm the dish, so the result is as accurate as possible. Any suggestions here if this would be the right way? Would love to hear your thoughts, and if this feels like something that would with your meal prep. Just sharing something I built because I wanted it to exist and solve a problem. June 16, 2025 at 12:23AM
Show HN: ZeroConfigDNLA – Easy to run media server in Python The goal was to be able to serve videos from my laptop in one command. Give it a go and let me know if it works for you! If you run into issues, please provide log output and the source and destination device info (make/model/etc) https://ift.tt/nx6udy1 June 15, 2025 at 03:46AM
Show HN: AnyCrawl v0.0.1-alpha.5 – custom user-agent and richer scraping API ## [0.0.1-alpha.5] - 2025-06-14 ### Added - Integrated AWS S3 storage support with new `S3` class and environment variables for seamless file uploads and retrievals. - Introduced `FileController` for serving files from S3 or local storage with robust path validation and error handling. - Added multiple content transformers (Screenshot, `HTMLTransformer`) improving HTML/Markdown extraction and screenshot generation. - Extended scraping capabilities with new options: output `formats`, `timeout`, tag filtering, `wait_for`, retry strategy, viewport configuration, and custom user-agent support. - Added Safe Search parameter to `SearchSchema` for filtered search results. - Refactored engine architecture with a factory pattern and new core modules for configuration validation, data extraction, and job management. - Implemented graceful shutdown handling for the API server and improved logging for uncaught exceptions / unhandled rejections. - Added Jest configuration for API and library packages with ESM support and updated test scripts. - Updated CI workflows to publish Docker images on version tags. - Expanded README with detailed environment variable descriptions and API usage examples. ### Changed - Refined error handling in `ScrapeController` and `JobManager`; failure responses now include structured error objects and HTTP status codes. - Enhanced `BaseEngine` with explicit HTTP error checks and resilience improvements. - Updated OpenAPI documentation to reflect new scraping parameters and error formats. - Migrated key-value store name to environment configuration for greater flexibility. - Enhanced per-request credit tracking in `ScrapeController` and enhanced logging middleware to include credit usage. ### Fixed - Improved job failure messages to include detailed error data, ensuring clearer debugging information. - Minor documentation corrections and clarifications. https://ift.tt/krvQ9nj June 14, 2025 at 11:18PM
Show HN: I built a Mac app to restore Dock-click minimize and avoid tiny buttons Hey HN, I'm the developer behind Click2Minimize. This app is my personal fix for two long-standing frustrations with the macOS interface. First, I wanted to restore Dock-click minimize. On other operating systems, I was used to clicking an app's icon to minimize its window—a simple, fast toggle. On a Mac, that second click does nothing, which always felt like a dead end in my workflow. Second, I was tired of having to deal with the tiny buttons. So much of window management—minimizing, maximizing, arranging—forces you to stop what you're doing, carefully aim your cursor at one of three small dots, and click. It's a constant micro-interruption. The Solution: A Fluid, Mouse-First Approach ----------------------------------------------------- Click2Minimize is a lightweight, native utility that turns your entire window title bar into a powerful gesture area. The goal is to let you manage your workspace without ever needing to aim for those little dots. * Consistent Dock Behavior: Click on Dock icon to minimize/hide the app. * Minimize Window Under Mouse: Simply hold down left mouse button and click the right one, or double-click the right button. * Maximize Window Under Mouse: Simply hold down right mouse button and click the left one, or double-click the notch area. * Snap Window to Left/Right: Simply hold down right button and rock the scroll wheel, or use fn key while swipe on trackpad. * Restore Window Size & Position: Holde down right button and click middle button, or user fn key with right-click on trackpad. * And many other useful gestures, such as the App Switcher and changing workspaces, were also included. Most importantly, it handled macOS full-screen mode smoothly and no longer felt intrusive. It is designed to resemble a missing feature of the operating system, with all gestures being highly intuitive, especially when using a mouse, as there is no need to remember keyboard shortcuts or bring the window to the front. Feedback, Discount & Free Licenses: ---------------------------------------- I'm here all day and would love to hear your thoughts. I genuinely want to make this app better, and the HN community's feedback is invaluable. Furthermore, I'll be sending a completely free license to the commenters with the most thoughtful feedback, bug reports, or feature suggestions I see. You're not just buying an app; you're helping to shape it. Link: https://ift.tt/yed0qGn https://ift.tt/yed0qGn June 14, 2025 at 11:51PM
Show HN: Shelly, terminal assistant that translates natural language into shell Describe what you want in plain English, and Shelly will figure out the right commands, explain what they do, and run them for you, with guardrails to ensure that you only run commands you feel safe running. https://ift.tt/jN4uGtH June 14, 2025 at 04:13AM
Show HN: Gem and I built an open-source app to learn Japanese I've been fascinated by the Japanese language and culture for a while now, and I wanted to create a simple, no-fuss way for beginners to get started. So, I built *[Nihongo]( https://nihongo.site/ )*, a free and open-source web app designed to help you learn the fundamentals of Japanese in about a month. The name of the app, Nihongo (日本語), is the Japanese word for the "Japanese language." You can check it out here: *[ https://nihongo.site/ ]( https://nihongo.site/ )* And for those who like to tinker, the code is available on GitHub: *[ https://ift.tt/lvwD1PA ]( https://ift.tt/lvwD1PA )* The "learn in 30 days" idea isn't about achieving fluency in a month, which we all know is impossible. Instead, the goal is to provide a structured and manageable learning path that covers the essential building blocks of the language in a short period. I wanted to create something that feels less intimidating than many comprehensive (and often expensive) resources out there. *What the app covers:* The app is structured into a series of lessons that you can follow at your own pace. It starts with the absolute basics and gradually introduces more complex concepts: * *The Japanese Writing Systems:* Detailed lessons on Hiragana and Katakana, the two phonetic scripts that are the foundation of written Japanese. * *Essential Grammar:* I've focused on the core grammatical structures you need to start forming your own sentences. * *Core Vocabulary:* You'll learn a curated list of high-frequency words that are immediately useful in everyday conversation. * *Practical Phrases:* The app includes common greetings and phrases that you can start using right away. *Why I built this:* I started building this project while testing the latest Gemini 2.5 models on Google AI Studio, and with the Code assistant and Cloud Run I was able to get it to production in less than 3 hours. This as a personal project to solidify my own understanding of Japanese and to build something useful for others who are just starting their learning journey. I'm a big believer in the power of open-source and wanted to create a resource that is accessible to everyone. This is very much a passion project, and I'm still actively working on it. I'd love to get your feedback, suggestions, and of course, any contributions on GitHub are more than welcome. Let me know what you think! I'm here to answer any questions you might have. https://nihongo.site June 14, 2025 at 04:34AM
Show HN: StellarSnap – Explore NASA APODs, simulate orbits, learn astronomy I built StellarSnap as a calm, ad-free space to explore NASA’s Astronomy Picture of the Day (APOD) and learn astronomy along the way. What it includes: - A clean APOD archive browser with a Random APOD button - A growing Glossary with term highlighting across the site - A 2D Orbit Simulator where you can test satellite motion with real physics - A deeper Encyclopedia, still early, but expanding - Subtle touches like “see past APODs using this term” - And more to come It’s entirely ad-free, cookie-free, and not affiliated with NASA, but I was honored to have StellarSnap mentioned on the official APOD About page by Professor Robert Nemiroff: https://ift.tt/DSXJOfx Always open to ideas, critiques, or ways to make it better. https://ift.tt/A8lEqcr June 13, 2025 at 10:32PM
Show HN: Dead simple clock for hidden menubar users I love keeping my menu bar hidden for a cleaner, distraction-free workspace. But constantly moving my cursor to the top just to check the time got annoying. IYKYK. So I built Corner Time - a minimal app that displays the current time in a carefully positioned screen corner, gives you instant time access while keeping your menu bar hidden. Quite simple, but it's genuinely improved my daily workflow. Features: • Always-visible time display • Customizable time format • Customizable font style I've been dogfooding this for weeks and it's become essential to my setup. With more Mac users embracing hidden menu bars (especially since recent macOS updates), figured others might find it useful too. Currently free on the Mac App Store - would love feedback from fellow hidden menu bar enthusiasts! https://ift.tt/HWt7LRs June 13, 2025 at 11:27PM
Show HN: I rebuilt the recruitment process from the ground up Hi HN Community, Recruitment software is everywhere. The market seems saturated. Every other day there’s a new ATS or “all-in-one” platform promising to fix hiring. But let’s be real — recruiting still sucks. Why? Because most tools are just reskinned versions of the same broken process: resume parsing, email campaigns, messy workflows, and outdated data. Some throw in a ChatGPT prompt here and there and call it “AI-powered.” But if we’re still stuck in the same flawed flow, it doesn't matter how modern the UI is. I’ve felt this pain personally — both as a recruiter and a job seeker. That's why I built Chronoflow — not just another ATS, but a reimagined recruitment system that actually works. --What makes it different: No resume parsing. No data entry. Candidate pools build themselves as soon as someone accepts your job invite. --No email campaigns. You already have the latest candidate data, and the platform shows you exactly who to engage. --No endless back-and-forth. Job invites include everything — replacing pre-screening calls and endless follow-ups. --Candidates get live updates and AI-generated feedback if rejected — improving their experience and keeping your brand strong. --Recruiters focus on decision making and building relationships, which is important for business development. Chronoflow is built for people who are tired of trying “yet another ATS” that solves 10% of the problem. If you're curious to see what a rebuilt hiring cycle looks like that drastically reduces time to hire and on top of that makes recruitment transparent and enjoyable, I'd love for you to give Chronoflow a try. Happy to answer any questions. And if you've worked in recruiting, would love to hear what frustrates you most — maybe I can solve that too. Thank You https://chronoflow.ai/ June 13, 2025 at 12:34AM
Show HN: ChatToSTL – AI text-to-CAD for 3D printing Hey HN, I'm a beginner at CAD so I built an app that does it for me ;) Describe a part and ChatToSTL writes the OpenSCAD code, shows a live render with size sliders, then exports the STL/3MF file. Because the output is parametric, it's easy to modify (unlike mesh models like Shap-E or DreamFusion). Try it (needs your own OpenAI key): https://ift.tt/mdhnzBa How it works: Text prompt → o4-mini generates OpenSCAD code → live render + sliders → refine in chat → export. Examples & Code: * Walkthrough + real prints (bowl, hook, box, door stop): https://ift.tt/pKnwtYL... * 90-sec demo: https://www.youtube.com/watch?v=ZK_IDaNn1Mk * MIT repo: https://ift.tt/8XlGgmP Current limitations (it's not replacing Fusion 360 anytime soon): - Simple shapes only. Even a mug can end up with a misplaced handle - Works best with CAD-style language ("extrude 5mm") - AI can't see the render, so no self-correction yet I'm particularly interested in feedback on improving the 3D generation quality: should I add vision feedback so that it can self critique? use CADQuery instead of OpenSCAD? use a different model? Thanks! Nico https://ift.tt/mdhnzBa June 12, 2025 at 11:28PM
Show HN: AuraCoder – Gen AI Learning Platform Hey HN, I've been pouring my time into this side project and I think I finally got an MVP up! I'm really excited about it. I'm really passionate about combining LLMs and learning. It seems like one of the best firsts for the tech. And so I built a site where all the content is generated. As I've been building, I'm always torn between building something more general purpose where you can learn anything vs building something targeted where the generation can be more tailored. Currently, its the latter so the site is focused on data structures and algorithms. That's something I've ground out recently so just familiar with what good content might look like and it was helpful in getting the prompt engineering to generate decent content. The site can generate both Lessons and Challenges. And they are a bit tailored to you. You can set settings about what kind of preferences you have. Tone of voice, depth, even an open text that gets feed into the prompt. I tried "Include a cat joke in every lesson" and I thought that was pretty entertaining It also takes into account your current skill level on different concepts. But I also think I need to lean in more on the customization. That seems to be the biggest way AI generated content can differentiate. I think its been hard to generate content that's really as good as human expert generated stuff, but it can be tailored to the user. So really interested in ideas in that vein. And in general, any advice is greatly welcomed. Also of course willing to AMA. Happy to answer any questions about the tech stack, the apps architecture, etc Sorry the site requires sign-up. I've thought about allowing anonymous users, but haven't implemented that yet. However, the site is free, and I'm not even doing any kind of email verification. So I won't judge you if you go with "some-fake-email@example.com" Hope your day is going well and all the best! https://auracoder.com/ June 12, 2025 at 07:31AM
Show HN: Eyesite - experimental website combining computer vision and web design I wanted Apple Vision Pros, but I don’t have $3,500 in my back pocket. So I made Apple Vision Pros at home. This was just a fun little project I made. Currently, the website doesn't work on screens less than 1200x728 (Sorry mobile users!) It also might struggle on lower end devices. For best results, have a webcam pointing right at you. I tested my website with a MacBook camera. Any comments, questions, or suggestions are greatly appreciated! blog: https://ift.tt/es7chML check it out: https://ift.tt/j4LSfHB github: https://ift.tt/nqDRPQz https://ift.tt/es7chML June 12, 2025 at 06:07AM
Show HN: The Roman Industrial Revolution that could have been This is a proof-of-concept comic book that asks: What if knowledge from 2025 reached Rome and kicked off an industrial revolution? The story follows two voices: - Ulysses, a present-day archaeologist who finds a glowing slate in the dig site. - Marcus, an educated household slave in 79 AD who replies on that slate. Why I’m posting: I’d love narrative feedback. – Does the story make sense? – Are Ulysses and Marcus believable? – Which directions would you explore next (politics, tech, moral fallout)? What’s live today - First issue, 25 rough pages. - No paywall; just a PDF. Next steps Regular releases toward a 8 or 10 issues collection. I’ll revise based on your critiques and wild speculations. Grateful for any thoughts on pacing, historical plausibility, or character depth. Thanks for reading! https://ift.tt/EXsuBFO June 12, 2025 at 05:21AM
Show HN: I made a 3D printed VTOL drone I made this 130 mile capable VTOL drone in only 90 days. It can fly for 3 hours on a single charge. That would make it one of the longest range and endurance 3D printed VTOLs in the world. This is the thing I'm most proud of building to date! Before this project, I was a total CAD, 3D printing and aerodynamic modeling beginner. I had only built and flown one VTOL before. SPECS Wingspan: 3.9 ft (1200 mm) Length: 2.5 ft (770 mm) Weight: 5.6 lb (2.55kg) Airframe: foaming PLA (Bambu PLA-Aero) and PETG structural parts printed on A1 printer, CFRP booms and spars Battery: Li-ion silicon anode Amprius SA08 cells, 6s2p pack by Upgrade Energy Motors: 2807 AOS for lift and cruise (unoptimized) Lifting ESCs: 4 in 1 Holybro Tekko32 F4 45A Cruise ESC: Flycolor Raptor 5 45A Lifting and cruise props: 7042 Gemfan (unoptimized) Flight controller: Speedybee F405 Wing GPS: M10 Firmware: Ardupilot 4.6.0 --- This video edit ended up shorter than I planned. Being my first Youtube video with significant post production effort, I underestimated the work required to make a longer in-depth video with voiceover, edited footage, etc. https://ift.tt/S9sdLZt June 11, 2025 at 02:17AM
Show HN: A "Course" as an MCP Server We wanted to build a course for new Mastra devs to get started quickly. However, we knew videos would go out of date and be more difficult to maintain. We decided to launch our "course" as an MCP server. This way your coding agent actually teaches the course content to you and can help you write the code. We think this is a really interactive way to learn. Using an editor with MCP support (such as Cursor, Windsurf, or VSCode), your code agent will call the appropriate MCP tools which will return context for the agent. This context tries to instruct the agent that it should be teaching you the content, not just doing the work for you. The course is still pretty experimental and some models work better than others. Code is available in the Mastra Github repo in the mcp-docs-server package ( https://ift.tt/sBaGvTZ... ) https://ift.tt/CBpStuc June 11, 2025 at 02:06AM
Show HN: RenderDay: A GPU-only render farm for Blender Hey HN! I recently had to render my first longer Blender animation, and I found myself pretty frustrated with the existing render farms out there. Everything I tried was either buggy, overly complicated (I really don’t want to pick from a huge list of hardware options), or just *really* expensive. So, I did what anyone would do.. I built my own solution: https://renderday.com - a GPU-only render farm for Blender that’s super fast and dead simple to use. You just: 1. Upload your `.blend` file 2. Pick your settings 3. Get a price, pay, and render - done No subscriptions, no upfront costs, no contracts - just pay as you go. I pull in daily GPU prices from multiple providers (with a tiny margin to keep the lights on), so the pricing is transparent and competitive. Under the hood it's running on NVIDIA L40S GPUs (48GB RAM), with access to over 1,000 GPUs globally. Currently supports: * Blender 4.3 and 4.4 (can add more if needed) * Cycles and EEVEE * Real-time progress tracking with live preview frames * Full file encryption, auto-deletion after 30 days, no access/sharing --- But more importantly: I'd really appreciate your feedback. This started as a personal itch, but I want to build something genuinely useful for the Blender community - especially indie creators and small studios who can't afford big monthly plans or don't want to deal with complicated setup. - What do you wish render farms did better? - What features are missing for you right now? - Would you use something like this - and if not, why not? Would love to hear your thoughts - good or bad - so I can keep improving it. Thanks for reading! Sascha https://renderday.com June 10, 2025 at 02:24AM
Show HN: I made a mobile app that turns your step count into a race I just launched my first-ever mobile app. It’s called STEPRACERS, a game where you compete with friends by tracking your steps. The idea came from someone close to me who completely changed their life by focusing on their health. Every night, they’d send me their step count - a small, daily ritual that became a powerful reminder of progress. So I turned it into a game. It’s fun. It’s simple. And it might just push you to hit your 10k steps a day. https://ift.tt/Hw915fp June 10, 2025 at 02:28AM
Show HN: Munal OS: a graphical experimental OS with WASM sandboxing Hello HN! Showing off the first version of Munal OS, an experimental operating system I have been writing in Rust on and off for the past few years. https://ift.tt/1nA8xzy It's an unikernel design that is compiled as a single EFI binary and does not use virtual address spaces for process isolation. Instead, applications are compiled to WASM and run inside of an embedded WASM engine. Other features: * Fully graphical interface in HD resolution with mouse and keyboard support * Desktop shell with window manager and contextual radial menus * PCI and VirtIO drivers * Ethernet and TCP stack * Customizable UI toolkit providing various widgets, responsive layouts and flexible text rendering * Embedded selection of applications including: * A web browser supporting DNS, HTTPS and very basic HTML * A text editor * A Python terminal Checkout the README for the technical breakdown. Demo video: https://ift.tt/zuiN3UY https://ift.tt/1nA8xzy June 9, 2025 at 11:04PM
Show HN: CurveFit Pro – Advanced nonlinear curve fitting in the browser We are pleased to announce a new version of our curve fitting web app with a whole new backend, front end, and all of the bells and whistles. If you see anything left out, or that could be improved, please let us know!! We'd love to hear any feedback from this amazing community. https://ift.tt/1FSZmQw June 8, 2025 at 09:01PM
Show HN: Liven Beta – Context engine mapping codebase dependencies for LLM(SWE) Liven Beta scaffolds your codebase’s dependency graph—functions, classes, cross‑file & folder relations—and displays it instantly in the terminal. Think of it as a context engine for developers and LLMs. https://ift.tt/PoTZjOw June 8, 2025 at 10:42PM
Show HN: AStack – A composable framework for building AI applications AStack is a composable framework designed to simplify the development of AI applications through a "everything is a component" philosophy. It provides a zero-adaptation layer design that enables seamless integration between various AI models, tools, and custom business logic. AStack is an independent technical framework with its own architecture and ecosystem, built on top of Hlang - a highly semantic fourth-generation language (4GL) inspired by Flow-Based Programming paradigms. This foundation on Hlang, which is particularly well-suited for computational modeling and AI-generated code, is what gives AStack its power. The framework emphasizes minimalism and performance, allowing developers to create complex systems with minimal boilerplate code while maintaining complete technical autonomy. https://astack.tech https://ift.tt/zfp0xy5 June 8, 2025 at 08:51AM
Show HN: AI that extracts brand identity from websites to generate ads I built this because I kept procrastinating on creating ads for my projects. The technical challenge was interesting: how do you teach AI to extract "brand identity" from a website? Turns out websites are messy. Finding the actual logo vs random images, identifying brand colors vs generic link colors, understanding brand voice from homepage copy. The solution: Custom vision models + CSS parsing + GPT-4 for voice analysis. You paste a URL, it extracts brand elements, generates platform-specific ads. Not trying to "disrupt advertising" or anything dramatic. Just solving the specific problem of "I need a Facebook ad but Canva makes me want to cry." Built with Next.js, custom image processing pipeline, OpenAI API. The brand extraction accuracy is around 85% for well-structured sites, lower for sites that are... creative with their CSS. Happy to discuss the technical approach or share code snippets if anyone's curious about the brand extraction pipeline. https://board.ad https://www.board.ad June 8, 2025 at 10:45AM
Show HN: AI game animation sprite generator I tried to build AI game animation generator last year ( https://ift.tt/mUp49sH ), a lot of people were interested, but it failed, mainly because the technology was not good enough. 1 year passed, there were a lot of developments in video/image generation. I tried it again, I think it works super well now. Actually beyond my expectation. You can generate all kinds of game character animation sprites with only 1 image. 1, upload your image of your character 2, choose the action you want 3, generate! Support basic actions like Run, Jump, Punch and complicated ones like: Shoryuken, Spinning kick, etc. High quality sprite sheet will be directly generated to use in Unity and any game engine. If you are an indie game developer, you don't need to high an artist or animator to develop you game. For studios, it's 10x cost saving and 10x efficiency as no more creating animations for 100 NPCs 100 times. Please check it out, looking forward to your feedback! https://ift.tt/nsXUA1p June 7, 2025 at 01:00AM
Show HN: Lightweight Durable Workflows Built on Postgres Hi HN! This is Qian here with Peter (KraftyOne) and Jeremy (jedberg). We’re building DBOS, an open-source, lightweight durable workflows library that you can add to Python apps in just a few lines of code. It’s comparable to popular open-source workflow and queue libraries like Airflow and Celery, but more lightweight with a greater focus on reliability and automatically recovering from failures. Our goal in building DBOS is to make workflows lightweight and flexible so you can add them to your existing apps with minimal work. Everything you need to run durable workflows and queues is contained in this Python library. You don’t need to manage a separate workflow server: just install the library, connect it to a Postgres database (to store workflow/queue state) and you’re good to go. DBOS workflows make your program durable by checkpointing its state in Postgres. If your program ever fails, when it restarts all your workflows will automatically resume from the last completed step. You add durable workflows to your existing program by annotating ordinary functions as workflows and steps: from dbos import DBOS @DBOS.step() def step_one(): ... @DBOS.step() def step_two(): ... @DBOS.workflow() def workflow(): step_one() step_two() The workflow is just an ordinary Python function. You can call it any way you like–from a FastAPI handler, in response to events, wherever you’d normally call a function. We’ve just released DBOS Python 1.0. This enhances workflows with many powerful features we’ve been building over the last few months, including: - Durable queues. Postgres-backed queues with all the queuing features of BullMQ/Celery (concurrency limits, rate limits, timeouts, priority, deduplication, etc.). Plus, they integrate with durable workflows, so you can write a workflow that enqueues 1K tasks, waits for and processes their results, and automatically recovers from any interruption. - Programmatic workflow management. Your workflows are stored as rows in a Postgres table, so you have full programmatic control over them. Write scripts to query workflow executions, batch pause or resume workflows, or even restart failed workflows from a specific step. This makes it much easier to diagnose and recover from bugs and failures that affect thousands of workflows. - Full support for both sync and async Python–write your workflows and steps as code either synchronously or asynchronously, it all works out of the box. - Improved tooling, including dashboards, workflow graph visualization, workflow management via web UI, and more. We’d love to hear your feedback and hope you can try DBOS out! https://ift.tt/Aqeu94f June 6, 2025 at 10:39PM
Show HN: Run 30B model in 4GB Active Memory We have built fused operator kernels for structured contextual sparsity to avoid loading and computing activations with feed forward layer weights that eventually zero out by the activation. The result? We are seeing 5X faster MLP layer performance in transformers with 50% lesser memory consumption avoiding the sleeping nodes in every token prediction. For Llama 3.2, Feed forward layers accounted for 30% of total weights and forward pass computation resulting in 1.6-1.8x increase in throughput: Sparse LLaMA 3.2 3B vs LLaMA 3.2 3B (on HuggingFace Implementation): - Time to First Token (TTFT): 1.51× faster (1.209s → 0.803s) - Output Generation Speed: 1.79× faster (0.7 → 1.2 tokens/sec) - Total Throughput: 1.78× faster (0.7 → 1.3 tokens/sec) - Memory Usage: 26.4% reduction (6.125GB → 4.15GB) Find the operator kernels with differential weight caching open sourced at github.com/NimbleEdge/sparse_transformers. Lets get LLMs sprinting! https://ift.tt/9zebDhZ June 5, 2025 at 10:43PM
Show HN: Smart Palette – Reimagining AI Art Hi HN, I'm Nick. Over the last 6 months I’ve been building Smart Palette – a platform to help anyone create unique, wall-ready art without needing to be a prompt expert. I started this because I wanted to unleash creativity in anyone and help them bring their art ideas onto their wall through a simple, guided and inspiring process. Instead of figuring out the “right” words to use, you just select your interior design style, room, art style, theme, and color palette. You can simply describe what you want to see and add your desired colors — or let Smart Palette handle it for you. Smart Palette uses a streamlined UI that my backend then translates into optimized, detailed prompts. A lot of the work went into this "translation" layer to ensure optimal model selection, settings and generation techniques depending on the user’s creative context. It also has a full print-on-demand (UHD) integration including various cusotmization options and an art preview feature. This is an early version, and I'd be very grateful for any feedback you have on the concept, the UX, or any technical aspects. Happy to answer any questions! You can try it out with a free trial and generate your first artwork. Here’s a quick walkthrough: https://ift.tt/T1ynlGh?... https://ift.tt/axqXS95 June 5, 2025 at 12:58AM
Show HN: SQLxport – Export SQL Query Results to Parquet, CSV, and S3 Introducing sqlxport: export SQL query results to Parquet/CSV in seconds — and optionally upload to S3 or MinIO. Works with PostgreSQL & Redshift CLI-first, cloud-friendly Great for data pipelines and engineers https://ift.tt/y81SCW2 June 3, 2025 at 11:56PM
Show HN: I build one absurd web project every month I’ve been building absurd, mostly useless web projects for fun — and I publish one every month at absurd.website. These are deliberately non-functional, weird, sometimes funny, sometimes philosophical — and usually totally unnecessary. Some examples: Sexy Math — solve math problems to reveal erotic images. Trip to Mars — a real-time simulation that takes 7 months to finish. Add Luck to Your e-Store — add a waving cat widget to boost your conversion via superstition. Microtasks for Meatbags — the future: AI gives prompts, humans execute. Invisible Lingerie — it’s sexy. And invisible. Artist Death Tracker — art prices spike when artists die. We track that. Open Celebrity — one open-source face, shared by all. Together we make her famous. I just enjoy exploring what the web can be when it doesn’t try to be “useful”. Would love to hear what you think — and absurd ideas are always welcome. https://absurd.website June 3, 2025 at 01:22AM
Show HN: GoogLLM – Google search that returns Markdown instead of HTML As part of my bigger goal to make the web more agent-friendly, this weekend i decided to tackle google. The "AI-native" search APIs like Tavily and Exa exist, but they require setup and don't actually use Google's results. So I built something simple - a proxy that takes Google search URLs and returns the results as clean markdown instead of HTML. You literally just change "google.com" to "googllm.com" in any search URL. ```bash # Returns 500KB of HTML: curl " https://ift.tt/zILSo7W " # Returns clean markdown: curl " https://ift.tt/ny39hHi " ``` *What it does:* - Serves normal HTML to browsers (so humans can use it normally) - Returns markdown to everything else (curl, fetch, LLM agents) - Supports all Google search types: web, images, news, scholar, shopping, etc. - No auth needed for testing (10 requests/hour free) *Technical approach:* - Content negotiation based on Accept headers - Caches results to avoid hammering Google - Simple pricing: 0.5¢ per search after free tier I built this over a long weekend because I was tired of writing HTML parsers for every project. The whole thing is designed around a single principle: make Google search results consumable by LLMs without any complexity. *Questions for HN:* - Is this approach too simplistic? Should search APIs be more complex? - How do you currently handle search in your LLM applications? - Any concerns about the proxy approach vs. building from scratch? The llms.txt documentation is intentionally comprehensive (2500 tokens) so any LLM can understand and use it immediately. Live demo: https://googllm.com API docs: https://ift.tt/4BfkVPq Would love feedback on the approach and any edge cases I might have missed. https://googllm.com June 3, 2025 at 12:37AM
Show HN: You2Aanki – Turn Videos into Anki Vocabulary Flashcards Hey HN, this is my first product launch. I built You2Anki along my language learning journey to aid my vocabulary from any content I want. Most tools I tried weren’t particularly made for language acquisition. You2Anki was designed with that focus in mind. Simple, intuitive and distraction-free. I hope it helps you! https://you2anki.com/ June 2, 2025 at 01:02AM
Show HN: I compressed 10k PDFs into a 1.4GB video for LLM memory While building a Retrieval-Augmented Generation (RAG) system, I was frustrated by my vector database consuming 8GB RAM just to search my own PDFs. After incurring $150 in cloud costs, I had an unconventional idea: what if I encoded my documents into video frames? The concept sounded absurd—storing text in video? But modern video codecs have been optimized for compression over decades. So, I converted text into QR codes, then encoded those as video frames, letting H.264/H.265 handle the compression. The results were surprising. 10,000 PDFs compressed down to a 1.4GB video file. Search latency was around 900ms compared to Pinecone’s 820ms—about 10% slower. However, RAM usage dropped from over 8GB to just 200MB, and it operates entirely offline without API keys or monthly fees. Technically, each document chunk is encoded into QR codes, which become video frames. Video compression handles redundancy between similar documents effectively. Search works by decoding relevant frame ranges based on a lightweight index. You get a vector database that’s just a video file you can copy anywhere. GitHub: https://ift.tt/EKYrijP https://ift.tt/EKYrijP May 29, 2025 at 06:24PM
Show HN: Purpose Reminders – One simple, positive act emailed monthly to all Hi HN, I built Purpose Reminders ( https://ift.tt/kjl1fMR ). Our first monthly action – "Leave a positive review for a local business" – goes out June 1st (very soon!). The core idea: What if thousands of us did the same small, positive act each month? You get one email, choose to act or skip (no pressure), and then see the collective, anonymous impact. It's 100% free, built with Next.js/Supabase/Resend. My attempt at a simple way to foster some collective goodwill. What do you think of the concept? https://ift.tt/kjl1fMR June 1, 2025 at 03:21AM
Show HN: A site for YC rejection stories Got rejected from YC a few times, so I built a site to collect lessons, reflections, and what people would do differently next time. Thought it could be helpful since most founders apply more than once anyway :) Hope it's helpful: https://ift.tt/2Z9VTaE https://ift.tt/2Z9VTaE June 1, 2025 at 12:45AM
संदेश
