Show HN: 1.2 users a day to keep the 9–5 away https://ift.tt/RBeWXyn

Show HN: 1.2 users a day to keep the 9–5 away In my long career as an “almost digital entrepreneur” (a fancy way to say I’ve tried a thousand things online without making a single cent), I never really felt that “this is it, I’m so close, I’ll finally quit everything and update my passport: job title? SaaS founder.” (Small detail: I don’t even have a passport. But I like to imagine that if I did, I’d want something cooler than “unemployed creative” written on it). For years, I collected side projects, hobbies, half-dead MVPs, and random nonsense, all with the same ending: super hyped at the beginning, burned out in the middle, completely abandoned by the end. But a couple years ago, I decided to take things more seriously (well… I try). I started building SaaS products. Simple, fast stuff, nothing too fancy. And finally, after a long toxic relationship with perfectionism, I realized something super basic but actually powerful: I don’t need thousands of users. I just need 1.2 paying users a day. Literally. Not to get rich, no Lamborghinis parked outside (also, I live in an apartment with no garage), but enough to live well, keep building, and maybe say “this is my job” without looking down in shame. It’s part math, part mindset. Like they told us in the first year of computer science: big problems get solved by breaking them into smaller ones. 100 users a day? Anxiety. 1.2 users a day? I can breathe. So yeah, this is my new mantra: “1.2 a day to keep the office job away.” Let’s see where this road takes me https://ift.tt/aGse8h9 April 30, 2025 at 11:15PM

Show HN: Typeconf – Dynamic Configs in TypeScript https://ift.tt/ig7zFG3

Show HN: Typeconf – Dynamic Configs in TypeScript Hi! I’m the co-creator of Typeconf, an open-source tool that lets you create dynamic configs like feature flags, texts, dictionaries with features, system prompts in Typescript. My personal goal for quite a while was to bring good infra practices from big tech to open-source world. Typeconf takes care of the problem of safe deployment. Like if you have a big service and growing amount of users, you want to keep adding new stuff quickly, you don’t want to break the app, and you don’t want to spend all your time in QA. That’s where dynamic configs help! I can see that existing alternatives are kinda more focused on analytics and A/B testing, and they are not code-first and moreover, not typesafe. This gap is what we’re aiming to cover with Typeconf. https://ift.tt/gXkfnU6 April 30, 2025 at 11:06PM

Show HN: ART – a new open-source RL framework for training agents https://ift.tt/SWeXVPE

Show HN: ART – a new open-source RL framework for training agents Hey HN, I wanted to share a new project we've been working on for the last couple of months called ART ( https://ift.tt/OclsnNU ). ART is a new open-source framework for training agents using reinforcement learning (RL). RL allows you to train an agent to perform better at any task whose outcome can be measured and quantified. There are many excellent projects focused on training LLMs with RL, such as GRPOTrainer ( https://ift.tt/4a7ntoJ ) and verl ( https://ift.tt/aRiksVh ). We've used these frameworks extensively for customer-facing projects at OpenPipe, but grew frustrated with some key limitations: - Multi-turn workflows, where the agent calls a tool, gets a response, and calls another, are not well supported. This makes them a non-starter for any task that requires an agent to perform a sequence of actions. - Other frameworks typically have low GPU efficiency. They may require multiple H100 GPUs just to train a small 7B parameter model, and aren't able to keep the GPUs busy consistently during both the "rollout" and "training" phases of the training loop. - Existing frameworks are typically not a convenient shape for integrating with existing agentic codebases. Existing trainers expect you to call raw text completion endpoints, and don't automatically provide industry-standard chat completion APIs. ART is designed to address these limitations and make it easy to train high-quality agents. We've also shared many details and practical lessons learned is in this post, which walks through a demo of training an email research agent that outperforms o3 ( https://ift.tt/OLgbmJe ). You can also find out more about ART's architecture in our announcement post ( https://ift.tt/xtJK2Sj... ). Happy to answer any questions you have! https://ift.tt/OclsnNU April 30, 2025 at 10:35PM

Show HN: Aisir – AI models deliberate and critique each other like a council https://ift.tt/U40Mvro

Show HN: Aisir – AI models deliberate and critique each other like a council https://ift.tt/KMt20fL April 30, 2025 at 10:10PM

Show HN: Dish – A simple HTTP and TCP endpoint monitoring tool (written in Go) https://ift.tt/JV5LlUa

Show HN: Dish – A simple HTTP and TCP endpoint monitoring tool (written in Go) Last month we posted about dish and shared the repo. Since then we have been working on a blog post showing it off in a bit more detail. You can find it in the link! https://ift.tt/VwLOCAP April 29, 2025 at 11:28PM

Show HN: BitNote – Using browsers and blockchains to store secrets https://ift.tt/nwTBchb

Show HN: BitNote – Using browsers and blockchains to store secrets https://ift.tt/NmiDav4 April 30, 2025 at 12:38AM

Show HN: Beatsync – perfect audio sync across multiple devices https://ift.tt/e7qxXGZ

Show HN: Beatsync – perfect audio sync across multiple devices Hi HN! I made Beatsync, an open-source browser-based audio player that syncs audio with millisecond-level accuracy across many devices. Try it live right now: https://ift.tt/ecu4OnM The idea is that with no additional hardware, you can turn any group of devices into a full surround sound system. MacBook speakers are particularly good. Inspired by Network Time Protocol (NTP), I do clock synchronization over websockets and use the Web Audio API to keep audio latency under a few ms. You can also drag devices around a virtual grid to simulate spatial audio — it changes the volume of each device depending on its distance to a virtual listening source! I've been working on this project for the past couple of weeks. Would love to hear your thoughts and ideas! https://ift.tt/jzKOwrQ April 30, 2025 at 12:32AM

Show HN: Sim Studio – Open-Source Agent Workflow GUI https://ift.tt/FuVT4Zd

Show HN: Sim Studio – Open-Source Agent Workflow GUI Hi HN! We're Emir and Waleed, and we're building Sim Studio ( https://simstudio.ai ), an open-source drag and drop UI for building and managing multi-agent workflows as a directed graph. You can define how agents interact with each other, use tools, and handle complex logic like branching, loops, transformations, and conditional execution. Our repo is https://ift.tt/1Smoyx4 , docs are at https://ift.tt/YOMRIPp , and we have a demo here: https://youtu.be/JlCktXTY8sE?si=uBAf0x-EKxZmT9w4 Building reliable, multi-step agent systems with current frameworks often gets complicated fast. In OpenAI's 'practical guide to building agents', they claim that the non-declarative approach and single multi-step agents are the best path forward, but from experience and experimentation, we disagree. Debugging these implicit flows across multiple agent calls and tool uses is painful, and iterating on the logic or prompts becomes slow. We built Sim Studio because we believe defining the workflow explicitly and visually is the key to building more reliable and maintainable agentic applications. In Sim Studio, you design the entire architecture, comprising of agent blocks that have system prompts, a variety of models (hosted and local via ollama), tools with granular tool use control, and structured output. We have plenty of pre-built integrations that you can use as standalone blocks or as tools for your agents. The nodes are all connected with if/else conditional blocks, llm-based routing, loops, and branching logic for specialized agents. Also, the visual graph isn't just for prototyping and is actually executable. You can run simulations of the workflows 1, 10, 100 times to see how modifying any small system prompt change, underlying model, or tool call change change impacts the overall performance of the workflow. You can trigger the workflows manually, deploy as an API and interact via HTTP, or schedule the workflows to run periodically. They can also be set up to trigger on incoming webhooks and deployed as standalone chat instances that can be password or domain-protected. We have granular trace spans, logs, and observability built-in so you can easily compare and contrast performance across different model providers and tools. All of these things enable a tighter feedback loop and significantly faster iteration. So far, users have built deep research agents to detect application fraud, chatbots to interface with their internal HR documentation, and agents to automate communication between manufacturing facilities. Sim Studio is Apache 2.0 licensed, and fully open source. We're excited about bringing a visual, workflow-centric approach to agent development. We think it makes building robust, complex agentic workflows far more accessible and reliable. We'd love to hear the HN community's thoughts! https://ift.tt/1Smoyx4 April 28, 2025 at 11:14PM

Show HN: A pure WebGL image editor with filters, crop and perspective correction https://ift.tt/RHbvYoQ

Show HN: A pure WebGL image editor with filters, crop and perspective correction I'm working on a pure js webgl image editor with effects, filters, crop & perspective correction, etc. My goal is to give the community an opensource solution as unfortunately most comparable apps are closed sources. https://ift.tt/WaHbKdm to try it out ( https://ift.tt/cqzG16R ) https://ift.tt/cqzG16R April 28, 2025 at 11:10PM

Show HN: NanoAgent, zero-dependency 1k-LOC AI-agent runtime https://ift.tt/MxsezIu

Show HN: NanoAgent, zero-dependency 1k-LOC AI-agent runtime https://ift.tt/AUWnjZc April 28, 2025 at 11:01PM

Show HN: Web-eval-agent – Let the coding agent debug itself https://ift.tt/jJtqw6u

Show HN: Web-eval-agent – Let the coding agent debug itself Hey HN! We’ve been building an MCP server to help AI-assisted web app developers by using browser agents to test whether changes made by an AI inside an editor actually work. We've been testing it on scenarios like verifying new flows in a UI, or checking that sending a chat request triggers a response. The idea is to let your coding agent both code and evaluate if what it did was correct. Here’s a short demo with Cursor: https://www.youtube.com/watch?v=_AoQK-bwR0w When building apps, we found the hardest part of AI-assisted coding isn’t the coding—it’s tedious point-and-click testing to see if things work. We got tired of this loop: open the app, click through flows, stare at the network tab, copy console errors to the editor, repeat. It felt obvious this should be AI-assisted too. If you can vibe-code, you should be able to vibe-test! Some agents like Cline and Windsurf have browser integrations, but Cline’s (via Anthropic Computer Use) felt slow and only reported console logs, and Windsurf’s didn’t work reliably yet. We got so tired of manually testing that we decided to fix it. Our MCP server sits between your IDE agent (Cursor/Windsurf/Cline/Continue) and a Playwright-powered browser-use agent. It spins up the browser, navigates your app per instructions from the IDE agent, and sends back steps, console events, and network events so the IDE agent can assess the app’s state. We proxy Browser-use’s original Claude calls and swap in Gemini Flash 2.0, cutting latency from ~8s → ~3s per step. We also cap console/network logs at 10,000 characters to stay within context limits, and filter out irrelevant logs (e.g., noisy XHR requests). At the end, the browser agent outputs a summary like: Web Evaluation Report for http://localhost:5173 Task: delete an API key and evaluate UX Steps: Home → Login → API Keys → Create Key → Delete Key Flow tested successfully; UX had problems X, Y, Z... Console (8)... Network (13)... Timeline of events (57) … This gives the coding agent the ability to recognize the console and network errors, or any issues with clicking around, and have the coding agent fix them before returning back to the user. (There’s a longer example in the README at https://ift.tt/NjRlP1Z .) Try it in Cursor / Cline / Windsurf / Claude Desktop: (macOS/Linux): curl -LSf https://ift.tt/pFTaECV -o install.sh less -N install.sh # inspect if you’d like bash install.sh # installs uv + jq + Playwright + server # then in Cursor/Cline/Windsurf/Continue: craft a prompt using the web_eval_agent tool (For Windows, there’s a 4-line manual install in the README.) What we want to do next: pause/go for OAuth screens; save/load browser auth states; Playwright step recording for automated test creation and regression test creation; supporting Loveable / v0 / Bolt.new sites by offering a web version. We’d love to hear your feedback, especially if you’ve experienced the pain of having to manually test changes happening in your web apps after making changes from inside your IDE, or if you’ve tried any alternative MCP tools for this that have worked well. Try it out if you feel it’d be helpful for your workflow: https://ift.tt/NjRlP1Z . (note: the server hits our operative.sh proxy to cover Gemini tokens. The MCP server itself is OSS; Anthropic base-URL support is coming soon. Free tier included; heavy users can grab the $10 plan to offset our model bill.) Let us know what you think! Thanks for reading! https://ift.tt/NjRlP1Z April 28, 2025 at 10:36PM

Show HN: A crowd-data collection Project about Indoor Air Quality (CO2) https://ift.tt/OCEbI94

Show HN: A crowd-data collection Project about Indoor Air Quality (CO2) I built IndoorCo2map.com and the accompanying (open source, MIT licensed) App with the goal to create a large free dataset about CO2 levels in public accessible Buildings and public transport using portable CO2-Monitors. We spent about 90% of our lives in doors, yet there is barely any data about the air quality in enclosed areas. CO2 is in most locations a very good proxy for the amount of exhaled air in a room which correlates with infection risks. High CO2 levels also can increase the aerostability of viruses so high levels are there is also a direct effect of CO2 as well. Aside from infection risks, high CO2 levels also can decrease cognitive abilities (during exposure, not permanently) and cause dizziness and headaches etc. Most existing studies are small size and focus on homes, hospitals or schools. During the beta test the community already took more than 10000 measurements (each between 5 and 120 minutes long), which to my knowledge already makes it the largest dataset of its kind. Currently most users are from german speaking regions which is a result of me being german and my social graph being mostly german. The App does not require any user registration and works with most of the common portable CO2-Monitors (Aranet4, Airvalent, Inkbird-IAM-T1, Airspot Health). For some of them I had to reverse engineer the Bluetooth messages and to make things worse the Airvalent’s data isn’t byte aligned. The App is built using C# MAUI and cross-compiles to both android and Apple. I use it because C# is the Language I am most comfortable in but also because it can be deployed to local iPhones without having to own a Mac. The Backend is using a queue server, serverless functions (also C#) and a postgresql Database, all hosted on AWS. The website is using maplibre, deck.gl and chart.js - I have no clue about websites so just tried to keep things simple. Expanding to other Indoor Air Quality indicators like PM2.5 would be trivial but currently the amount of people having mobile sensors is too small to be worth the effort. https://ift.tt/liBMrqz April 27, 2025 at 10:25PM

Show HN: Mirror AI – LLM agent that takes action, not just chat https://ift.tt/37ekxcr

Show HN: Mirror AI – LLM agent that takes action, not just chat Hi HN, TL;DR Mirror AI is a cross-platform desktop action oriented LLM . Instead of just replying with text, it can run terminal commands, move files, hit APIs, send email/messages, schedule calendar events, query databases etc. - and chain those steps in one request. Why I built it Chatbots are great for answers but stop at the “OK, now go do it” moment. I was copy-pasting between ChatGPT and my shell/IDE/Gmail far too often, so on red-eye flights I hacked a bridge that lets an LLM execute tasks under a permission layer. Friends asked for it, feature-creep happened, and here we are. What it can do today - Files & Shell – read/write/rename/compress/execute files, run shell commands - Productivity – draft + send Gmail, create Google Calendar events, read/write Notion, post to Slack/Discord/WhatsApp (via Beeper) - Dev/Ops – query Postgres/Mongo, interact with GitHub, call AWS, curl any REST endpoint - Search/GenAI – web search (Perplexity), summarise PDFs, generate images/video, keep long-term memory - Extensible via MCP (Modular Command Protocol) to add your own skills - And a lot more Example prompt after you have added a few integrations (copy-paste): Search Reddit for discussion on LangChain streaming, summarise it, email me the summary, and block 30 min in Calendar tomorrow. OR Investigate my AWS account and tell me how to easily save money. Also check my MongoDB and make some proposition to optimize it. How it works - Desktop app (macOS, Windows, Linux) written in TS - No SaaS backend: everything runs locally; outbound calls only hit the model provider you configure (OpenAI, Claude, local Ollama, etc.) - Safety: every risky actions require approval from the user Status & roadmap - Free alpha download: https://themirrorai.com (bring your own API keys) - Planning to open-source core in the coming months after tidying up if there is enough interest - Up next: improved working memory, event-driven triggers for autonomous responses, and expanded integrations. Ask Kick the tires, try weird edge-cases, look for security holes, and tell me what hurts or what you wish it did. I will be in the thread checking the comments. Thanks! Sat https://themirrorai.com April 27, 2025 at 09:47PM

Show HN: Logchef – Schema-agnostic log viewer for ClickHouse https://ift.tt/7HOQWv8

Show HN: Logchef – Schema-agnostic log viewer for ClickHouse Hey HN! I’m Karan, creator of Logchef ( https://logchef.app ), an open-source log viewer built specifically for exploring logs stored in ClickHouse. This tool grew directly out of my $day job managing massive log volumes. Like many orgs, we migrated our log workloads to ClickHouse for its performance, but found the ecosystem lacked dedicated UI tooling for actually browsing and analyzing those logs effectively. We were using Metabase, and while great for general BI, it wasn't designed for log exploration workflows. Common pain points included: - Clunky Ad-hoc Querying: Writing/modifying raw ClickHouse SQL for quick checks was slow and error-prone, especially during incidents. - Disconnect Between Viz & Raw Logs: Visualizing trends (like error counts) then drilling down to the specific raw logs often required separate, complex queries and wrestling with row limits. The intuitive "slice-and-dice" was missing. - UI Friction: Simple things like selecting precise time ranges ("last 90 minutes"), easily viewing surrounding log context, or dealing with truncated columns added unnecessary friction. Debugging sessions were taking longer than they should. So, over the last 3-4 months, I built Logchef to scratch this itch. Logchef's Core Ideas: - Purpose-Built for ClickHouse Logs: Designed from the ground up for the specific task of log exploration on top of ClickHouse, focusing on speed and intuitive workflows. - Schema-Agnostic: Logchef doesn't force OTEL or any other schema. Connect it directly to your existing ClickHouse log tables (it just needs a timestamp column). Bring your own schema! - Focus on Viewing/Querying: Logchef intentionally doesn't handle log collection/ingestion. It complements great tools like Vector, Fluentbit, etc., by focusing purely on the exploration layer once data is in ClickHouse. - Simple Search Syntax: Includes a simple query syntax (e.g., `status=200 and path~"/api/"`) that translates to efficient ClickHouse SQL behind the scenes, integrated with the Monaco editor. Tech Stack: Go backend, SQLite for metadata, Vue.js + shadcn/ui + Tailwind CSS frontend. You can try a live public demo here: https://ift.tt/PWxw1gr (It's pre-populated with sample data using Vector, so you can dive right in. Uses Dex for OIDC auth - creds are on the login page). What's Next & Getting Involved: Logchef is already used internally at Zerodha, and I'm driving towards v1.0 this year. The roadmap includes features like Alerting, Live Tail Logs, and Enhanced Dashboarding. It's open source (AGPLv3), and I'd love to get more eyes on it and build a community. Check out the repo: https://ift.tt/39EBlgD I’d love to hear your feedback, whether positive or negative. Please open issues on GitHub with suggestions or bug reports! Thanks so much, HN! https://ift.tt/39EBlgD April 27, 2025 at 10:15PM

Show HN: Install CLI Apps via Plain HTTP – No Docker, No Binaries, Just Curl https://ift.tt/qL4Golj

Show HN: Install CLI Apps via Plain HTTP – No Docker, No Binaries, Just Curl Tired of bloated installers and complex DevOps pipelines? I built PPORT — a terminal-based messenger — to demo a crazy simple idea: 1. Instant CLI delivery over HTTP 2. Just curl or irm, nothing else 3. TypeScript on the fly via Deno 4. Live deployment without Docker or builds How it works: Visit https://pport.top Run one command (curl -fsSL pport.top | sh) PPORT streams scripts and source files dynamically based on your client (curl, browser, Deno) No packaging. No compiling. No friction. Source on GitHub: https://ift.tt/k4JRHMO Curious what else could be built with this approach? Would love to hear your ideas. https://pport.top April 26, 2025 at 11:51PM

Show HN: Rocal UI – A simple template engine with Rust https://ift.tt/qHIdoXu

Show HN: Rocal UI – A simple template engine with Rust https://ift.tt/oFa1bPR April 27, 2025 at 01:26AM

Show HN: AgenticSeek – Self-hosted Manus alternative https://ift.tt/WbYqVP7

Show HN: AgenticSeek – Self-hosted Manus alternative I’ve spent the last two months building AgenticSeek, a privacy-focused alternative to cloud-based AI tools like ManusAI. It runs entirely on your machine—no API calls, no data leaks. Why AgenticSeek? Optimized for local LLMs (developed mostly on an RTX 3060 running deepseek r1 14b). Truly private: All components (TTS, STT, planner) run locally. More responsive than alternatives (we respond fast to issues + active Discord). Designed to be fun—think JARVIS-like voice control, multi-agent workflows, and a slick web UI. Current Features: Web browsing (research + form filling), code write/fix, file management/search. Planning capabilites to use multiple agents for complex task. Is it stable? Prototype-stage—great for tinkerers. Hoping to get feedback! https://ift.tt/0AmGIYF April 27, 2025 at 12:23AM

Show HN: I build a Fantasy NHL app in 3 days with Claude AI https://ift.tt/fau6RTI

Show HN: I build a Fantasy NHL app in 3 days with Claude AI I am not from North America, and moved there a couple years ago. After being invited to the first fantasy league and seeing the platform they were using, I couldn't bring myself to check it every day. An excel sheet would have been more appealing to use than this. So I decided to create my own. But on the side? This would take ages. I am very confident with Rust and JavaScript, but it still takes ages to build something like that. Good prompts, and a few long nights later, and I could create something fun and easy to use. Have a look: https://ift.tt/Z0yFhoE The frontend is in React, the backend in Rust. I deployed both via fly.io. It was so simple and fast, it was shocking. After it was deployed, I took 2-3 days to refactor everything and made it neat and tidy, so I could possibly open it up for general use with accounts etc. The first time I built something with the help of AI not just for me. The biggest help was certainly the UI and styling. I was never good at that. The first draft looked rough. I fed it some screenshots from NHL.com and told the model to "make it look nicer". What you see is I think the 20th iteration of the app, slowly improving, fixing bugs etc. But 90% with the help of AI. I could have done 98% myself (except the styling part). And after each working iteration, I spend quite some time cleaning up so I can build on top of that. The beauty? The cleaner, more modular code base saved tokens and made it easier for the model to refactor and understand. Strictly typed (TypeScritp in the frontend, Rust in the backend) helped as well! April 27, 2025 at 12:04AM

Show HN: Claude Code with GUI and Block Based Prompt Editor (MIT) https://ift.tt/7qIAbCZ

Show HN: Claude Code with GUI and Block Based Prompt Editor (MIT) https://ift.tt/qsE2NW9 April 25, 2025 at 11:58PM

Show HN: KRDS – Korea Government Design System in React and MUI https://ift.tt/g0L7y2a

Show HN: KRDS – Korea Government Design System in React and MUI I built an implementation of the Korea Design System (krds.go.kr) using React and MUI. - Storybook: https://ift.tt/tQiO3Cp - NPM: https://ift.tt/C45lF0W - GitHub: https://ift.tt/dOqjGFK - Figma: https://ift.tt/HYxcta2 To support developers working on public sector UIs, or anyone exploring frontend architecture and design systems in real-world projects. Current limitations: - Icons are from @mui/icons-material for now - Design tokens are static and will be refactored to improve developer experience. Feedback & Contributions: - Open to evolving this further with community input. - PRs and issues are always welcome, and I'd consider moving it to an organization if there's enough interest. https://ift.tt/dOqjGFK April 25, 2025 at 09:32PM

Show HN: Magnitude – open-source, AI-native test framework for web apps https://ift.tt/HfWNjTh

Show HN: Magnitude – open-source, AI-native test framework for web apps Hey HN, Anders and Tom here - we’ve been building an end-to-end testing framework powered by visual LLM agents to replace traditional web testing. We know there's a lot of noise about different browser agents. If you've tried any of them, you know they're slow, expensive, and inconsistent. That's why we built an agent specifically for running test cases and optimized it just for that: - Pure vision instead of error prone "set-of-marks" system (the colorful boxes you see in browser-use for example) - Use tiny VLM (Moondream) instead of OpenAI/Anthropic computer use for dramatically faster and cheaper execution - Use two agents: one for planning and adapting test cases and one for executing them quickly and consistently. The idea is the planner builds up a general plan which the executor runs. We can save this plan and re-run it with only the executor for quick, cheap, and consistent runs. When something goes wrong, it can kick back out to the planner agent and re-adjust the test. It’s completely open source. Would love to have more people try it out and tell us how we can make it great. Repo: https://ift.tt/2PlanEV Discord here if you wanna stay up to date or ask us anything: https://ift.tt/5a1btwx https://ift.tt/2PlanEV April 26, 2025 at 12:00AM

Show HN: We made a blazing-fast, open-source GitHub front end https://ift.tt/zQxYtlm

Show HN: We made a blazing-fast, open-source GitHub front end https://gitfaster.dev/ April 24, 2025 at 11:36PM

Show HN: Chat with AI copies of all HN users, like sama, dang, and even yourself https://ift.tt/PVN1Qa4

Show HN: Chat with AI copies of all HN users, like sama, dang, and even yourself Hey HN, I'm Chris. I've been working on creating personalized AI chatbots by grounding them in real conversations. Feeling inspired by antirez's recent post ("Reproducing Hacker News writing style fingerprinting", https://ift.tt/rAfMz2K ), I wondered: could I use the HN comment data referenced in that post to create virtual versions of... every user on Hacker News? A ~week later, here we are: https://ift.tt/WT4vfnX The database contains every user who has commented at least 200 words, and when you search for a user, we'll process their history in realtime to create a personalized AI chatbot of them on-the-fly (it'll take a few minutes). There's no login required, and conversations aren't recorded or stored anywhere, nor are they trained on. It surprised me at how fun it actually is. Give it a shot, would love to know what y'all think-- https://ift.tt/WT4vfnX April 24, 2025 at 08:43PM

Show HN: Moose – OSS framework to build analytical back ends with ClickHouse https://ift.tt/SQkVeJH

Show HN: Moose – OSS framework to build analytical back ends with ClickHouse https://ift.tt/feKwPda April 23, 2025 at 11:41PM

Show HN: Bloggr – a Next.js blogging engine (MIT) https://ift.tt/7gO1bwA

Show HN: Bloggr – a Next.js blogging engine (MIT) Hi HN! I’m excited to open-source Bloggr, a lightweight, Next.js-first blogging engine backed by Supabase (by default). # Repo: https://ift.tt/OztGFlk # Blog demo: https://bloggr.dev/blog Key Features: - Modern Next.js 15.3 App Router foundation - Secure admin area (CRUD) with Supabase auth & middleware protection - Tiptap rich text editor with image uploads (to Supabase Storage Bucket) - Blog index pagination - Per-post SEO tags TODO: - Add themes - Add blog posts and blog index layout customisation - Add other data layers Feedback Welcome: What features are crucial for your blogging needs? What other data layers (PostgreSQL direct, MongoDB) would you like adapters for? P.S.: Node.js version is coming soon https://bloggr.dev April 24, 2025 at 12:01AM

Show HN: MCP Manager – Open-Source CLI to Manage Your MCP Servers https://ift.tt/IJS6gGa

Show HN: MCP Manager – Open-Source CLI to Manage Your MCP Servers Hi HN! I'm Noah, an ordinary guy that got really interested in MCP about a month ago when I downloaded Cursor. It's been super exciting to see companies around the world launch their own MCP servers empowering AI to do more. Unfortunately, I got a bit overwhelmed setting up MCP servers across my clients like Claude Desktop, Cursor, and Claude Code, so I figured I'd take a shot at building an open-source tool to help others and myself! So yeah, I built a CLI tool that enables you to quickly get MCP servers from the Web integrated with your clients like Cursor and Claude Desktop. You can use the built-in MCP servers I've defined like Playwright, Github, etc and add your own configurations! Feel free to support me too. PRs are welcome! Thanks for reading this and have a great day! https://ift.tt/inS4RUY April 24, 2025 at 12:00AM

Show HN: Advanced-Alchemy – A framework agnostic library for SQLAlchemy https://ift.tt/HZKNU42

Show HN: Advanced-Alchemy – A framework agnostic library for SQLAlchemy Hello HN! Advanced Alchemy is an optimized companion library for SQLAlchemy, designed to supercharge your database models with powerful tooling for migrations, asynchronous support, lifecycle hook and more. You can find the repository and documentation here: - GitHub Repository: https://ift.tt/S7IuZpD - Official Documentation : https://ift.tt/AaoBLW5 Advanced-Alchemy extends SQLAlchemy with productivity-enhancing features, while keeping full compatibility with the ecosystem you already know. At its core, Advanced-Alchemy offers: - Sync and async repositories, featuring common CRUD and highly optimized bulk operations - Integration with major web frameworks including Litestar, Starlette, FastAPI, Sanic - Custom-built alembic configuration and CLI with optional framework integration - Utility base classes with audit columns, primary keys and utility functions - Built in File Object data type for storing objects: - Unified interface for various storage backends fsspec and obstore - Optional lifecycle event hooks integrated with SQLAlchemy's event system to automatically save and delete files as records are inserted, updated, or deleted. - Optimized JSON types including a custom JSON type for Oracle - Integrated support for UUID6 and UUID7 using uuid-utils - Integrated support for Nano ID using fastnanoid - Pre-configured base classes with audit columns UUID or Big Integer primary keys and a sentinel column. - Synchronous and asynchronous repositories featuring: - Common CRUD operations for SQLAlchemy models - Bulk inserts, updates, upserts, and deletes with dialect-specific enhancements - Integrated counts, pagination, sorting, filtering with LIKE, IN, and dates before and/or after. - Tested support for multiple database backends including: - SQLite via aiosqlite or sqlite - Postgres via asyncpg or psycopg3 (async or sync) - MySQL via asyncmy - Oracle via oracledb (async or sync) (tested on 18c and 23c) - Google Spanner via spanner-sqlalchemy - DuckDB via duckdb_engine - Microsoft SQL Server via pyodbc or aioodbc - CockroachDB via sqlalchemy-cockroachdb (async or sync) The framework is designed to be lightweight yet powerful, with a clean API that makes it easy to integrate into existing projects. You can find a full example using FastAPI there: https://ift.tt/MAUFIgX... There are custom datatypes, a service and repository (including optimized bulk operations), and native integration with Flask, FastAPI, Starlette, Litestar and Sanic. Feedback and enhancements are always welcomed! We have an active discord community, so if you don't get a response on an issue or would like to chat directly with the dev team, please reach out. The library: https://ift.tt/S7IuZpD https://ift.tt/S7IuZpD April 23, 2025 at 08:17PM

Show HN: I open-sourced my AI toy company that runs on ESP32 and OpenAI realtime https://ift.tt/LQM4u5S

Show HN: I open-sourced my AI toy company that runs on ESP32 and OpenAI realtime Hi HN! Last year the project I launched here got a lot of good feedback on creating speech to speech AI on the ESP32. Recently I revamped the whole stack, iterated on that feedback and made our project fully open-source—all of the client, hardware, firmware code. This Github repo turns an ESP32-S3 into a realtime AI speech companion using the OpenAI Realtime API, Arduino WebSockets, Deno Edge Functions, and a full-stack web interface. You can talk to your own custom AI character, and it responds instantly. I couldn't find a resource that helped set up a reliable, secure websocket (WSS) AI speech to speech service. While there are several useful Text-To-Speech (TTS) and Speech-To-Text (STT) repos out there, I believe none gets Speech-To-Speech right. OpenAI launched an embedded-repo late last year which sets up WebRTC with ESP-IDF. However, it's not beginner friendly and doesn't have a server side component for business logic. This repo is an attempt at solving the above pains and creating a great speech to speech experience on Arduino with Secure Websockets using Edge Servers (with Deno/Supabase Edge Functions) for fast global connectivity and low latency. https://ift.tt/DHNfz8s April 22, 2025 at 09:10PM

Show HN: TableTennis, a new rubygem for printing stylish tables in your terminal https://ift.tt/q9lgD62

Show HN: TableTennis, a new rubygem for printing stylish tables in your terminal https://ift.tt/WvJZ7d3 April 23, 2025 at 12:00AM

Show HN: I built a Ruby gem that handles memoization with a ttl https://ift.tt/kobKcMn

Show HN: I built a Ruby gem that handles memoization with a ttl I built a Ruby gem for memoization with TTL + LRU cache. It’s thread-safe, and has been helpful in my own apps. Would love to get some feedback: https://ift.tt/EdLjJhY https://ift.tt/EdLjJhY April 22, 2025 at 11:51PM

Show HN: Morphik – Open-source RAG that understands PDF images, runs locally https://ift.tt/jXgQxnG

Show HN: Morphik – Open-source RAG that understands PDF images, runs locally Hey HN, we’re Adi and Arnav. A few months ago, we hit a wall trying to get LLMs to answer questions over research papers and instruction manuals. Everything worked fine, until the answer lived inside an image or diagram embedded in the PDF. Even GPT‑4o flubbed it (we recently tried O3 with the same, and surprisingly it flubbed it too). Naive RAG pipelines just pulled in some text chunks and ignored the rest. We took an invention disclosure PDF ( https://ift.tt/JWflUZH... ) containing an IRR‑vs‑frequency graph and asked GPT “From the graph, at what frequency is the IRR maximized?”. We originally tried this on gpt-4o, but while writing this used the new natively multimodal model o4‑mini‑high. After a 30‑second thinking pause, it asked for clarifications, then churned out buggy code, pulled data from the wrong page, and still couldn’t answer the question. We wrote up the full story with screenshots here: https://ift.tt/mi35sjC . We got frustrated enough to try fixing it ourselves. We built Morphik to do multimodal retrieval over documents like PDFs, where images and diagrams matter as much as the text. To do this, we use Colpali-style embeddings, which treat each document page as an image and generate multi-vector representations. These embeddings capture layout, typography, and visual context, allowing retrieval to get a whole table or schematic, not just nearby tokens. Along with vector search, this could now retrieve exact pages with relevant diagrams and pass them as images to the LLM to get relevant answers. It’s able to answer the question with an 8B llama 3.1 vision running locally! Early pharma testers hit our system with queries like "Which EGFR inhibitors at 50 mg showed ≥ 30% tumor reduction?" We correctly returned the right tables and plots, but still hit a bottleneck, we weren’t able to join the dots across multiple reports. So we built a knowledge graph: we tag entities in both text and images, normalize synonyms (Erlotinib → EGFR inhibitor), infer relations (e.g. administered_at, yields_reduction), and stitch everything into a graph. Now a single query could traverse that graph across documents and surface a coherent, cross‑document answer along with the correct pages as images. To illustrate that, and just for fun, we built a graph of 100 Paul Graham’s essays here: https://ift.tt/9aq4ELe You can search for various nodes, (eg. startup, sam altman, paul graham and see corresponding connections). In our system, we create graphs and store the relevant text chunks along with the entities, so on querying, we can extract the relevant entity, do a search on the graph and pull in the text chunks of all connected nodes, improving cross document queries. For longer or multi-turn queries, we added persistent KV caching, which stores intermediate key-value states from transformer attention layers. Instead of recomputing attention from scratch every time, we reuse prior layers, speeding up repeated queries and letting us handle much longer context windows. We’re open‑source under the MIT Expat license: https://ift.tt/nl76Pyz Would love to hear your RAG horror stories, what worked, what didn’t and any feedback on Morphik. We’re here for it. https://ift.tt/nl76Pyz April 22, 2025 at 11:18PM

Show HN: I built Run for Fun to stop me from doomscrolling until I exercise https://ift.tt/6veqdjY

Show HN: I built Run for Fun to stop me from doomscrolling until I exercise Some of you might know me as the developer of HACK, the iOS and Android apps for Hacker News. I spend a lot of time doomscrolling on Reddit, YouTube, and here at Hacker News. I also enjoy Powerlifting. So, I wondered whether I could use exercising as a way to reduce doomscrolling. So, I built Run for Fun. It helps you reduce screen addiction, end doomscrolling, practice self-discipline, and get healthy. You can privately and securely limit the usage of addictive social media apps and websites until you perform physical activity. When you walk, run, bike, climb stairs, and exercise, you gain screen time points to use those apps. You can control how easy or hard it is to gain those points. You can choose which weekdays and between what time period restrictions apply. You can snooze restrictions, choose how long to snooze for, or disable the snooze button altogether. I built it in Swift and according to Apple, is only 860 kB to download and 1.8 MB to install. I pride myself in keeping downloads small. It uses Apple's Screen Time and HealthKit APIs. The physical activity and app usage data stays private and is not transmitted to me, the developer. There are no ads and no subscriptions. Free version is limited to restricting 1 app, 1 website, 1 category and all-day everyday schedule. A one time only in app purchase gets you lifetime access to restricting unlimited apps, websites, and categories, restrictions scheduled on specific weekdays and time, and custom screen time points calculation and snooze times. Is this something you may find useful? https://ift.tt/9zgrBhO April 22, 2025 at 12:18AM

Show HN: Dia, an open-weights TTS model for generating realistic dialogue https://ift.tt/54tqYQE

Show HN: Dia, an open-weights TTS model for generating realistic dialogue https://ift.tt/yQxvtFW April 22, 2025 at 12:07AM

Show HN: Light like the Terminal – Meet GTK LLM Chat Front End https://ift.tt/YTEBF2G

Show HN: Light like the Terminal – Meet GTK LLM Chat Front End Author here. I wanted to keep my conversation with #Gemini about code handy while discussing something creative with #ChatGPT and using #DeepSeek in another window. I think it's a waste to have Electron apps and so wanted to chat with LLMs on my own terms. When I discovered the llm CLI tool I really wanted to have convenient and pretty looking access to my conversations, and so I wrote gtk-llm-chat - a plugin for llm that provides an applet and a simple window to interact with LLM models. Make sure you've configure llm first ( https://ift.tt/qCNIH6R ) I'd love to get feedback, PRs and who knows, perhaps a coffee! https://ift.tt/HTFDh2X https://ift.tt/RGSn4vT April 21, 2025 at 11:36PM

Show HN: Pad.ws – Whiteboard IDE is now open-source https://ift.tt/kClBGbt

Show HN: Pad.ws – Whiteboard IDE is now open-source Pad.ws is an infinite canvas that runs Linux, you can draw, write, open terminals or vscode to your very own ubuntu workspace. This uses Excalidraw for the canvas part while Coder manages the dev environments. We have a Public hosted instance at https://pad.ws if you want to try it out Disclaimer: this project is still very early, your feedback helps us shape it, so please let us hear your ideas! https://ift.tt/C7N9F38 April 21, 2025 at 11:20PM

Show HN: Tascli, a simple, local CLI task and record manager https://ift.tt/O36R8Ed

Show HN: Tascli, a simple, local CLI task and record manager A simple task and record manager in CLI, with entries stored in a local sqlite file. Treats task (things with a due date) and records (things with only event date) separately and able to manage both. https://ift.tt/McfJeU9 April 21, 2025 at 02:01AM

Show HN: npx -y @considered/harmful https://ift.tt/4wV0FIa

Show HN: npx -y @considered/harmful https://ift.tt/crd8NTJ April 21, 2025 at 12:07AM

Show HN: A tool that makes numbers pop, flip, shrink, or emoji-fy using Unicode https://ift.tt/fHdBZYs

Show HN: A tool that makes numbers pop, flip, shrink, or emoji-fy using Unicode https://ift.tt/8izMLTO April 20, 2025 at 11:46PM

Show HN: Learn bird calls with audio flashcards https://ift.tt/ArpXl0C

Show HN: Learn bird calls with audio flashcards https://ift.tt/ZgnopDk April 20, 2025 at 11:31PM

Show HN: I built an AI that turns GitHub codebases into easy tutorials https://ift.tt/1ICEcQf

Show HN: I built an AI that turns GitHub codebases into easy tutorials https://ift.tt/2KWv09E April 20, 2025 at 04:04AM

Show HN: Ibex – a cross-platform iOS backup decryption tool https://ift.tt/8Vx0UTn

Show HN: Ibex – a cross-platform iOS backup decryption tool ibex is a cross-platform tool designed for decrypting and extracting iOS backups. It provides forensic investigators, security researchers, and power users with the ability to access and analyze encrypted iOS backup data. It can be built and used on macOS, Linux, and Windows and is permitted to be used only with the explicit and informed consent of the backup data owner. Ibex was written in Go for straightforward compilation and to circumvent dependency issues and with the goal of enabling researchers and defenders assisting civil society victims of spyware and stalkerware Key Features - Decrypt encrypted iOS backups - Support for latest iOS versions - Cross-platform compatibility (macOS, Windows, Linux) - Automatic backup detection - Single file extraction based on filename match - Structured output organization - Detailed manifest parsing and extraction Basic Usage Examples # Run with automatic backup detection and interactive mode ibex # Specify just the backup path ibex -b /path/to/backup # Specify backup path and password ibex -b /path/to/backup -p "backup_password" # Specify custom output directory ibex -b /path/to/backup -p "backup_password" -o /path/to/output # Specify a single file for decryption and extraction ibex -b /path/to/backup -o /path/to/output --file sms.db # Specify relative path preserved output ibex -b /path/to/backup -o /path/to/output -r https://ift.tt/Pry5mFx April 20, 2025 at 12:40AM

Show HN: Web Video editor, 100% local, AI subtitle, auto cut based on volume https://ift.tt/JVpx9MD

Show HN: Web Video editor, 100% local, AI subtitle, auto cut based on volume I made this mostly to make content creation faster. When recording a youtube video/ podcast. there's alot of downtime (silence) while looking at scripts or getting the correct item to display. So I made a webapp to automatically cut and stich the video back together then I can further edit in davinci resolve. But I recently found out that this is crazy good for rewatching lectures or meeting recordings, No need to render, just drop the video in and watch it right away. 30min lecture can be watched in 12min! I even added auto generation of subtitles to make studying lecture /meeting better. Best of all everything runs locally, even the AI is running local models. Hope HN finds this useful like it did for me! https://ift.tt/wbt0B5C April 19, 2025 at 11:00PM

Show HN: I built a simple, fast transit app for the Bay Area https://ift.tt/5ZoxkXA

Show HN: I built a simple, fast transit app for the Bay Area Hey HN, I built Commuter because I was tired of switching between different apps to check arrival times for BART, Caltrain, Muni, ferries, and more. This app pulls directly from the official 511 API and aims to provide a fast, clean experience focused on real-time departures. There’s no account creation, it’s free to use, and it supports every major transit provider in the Bay Area—from Napa down to San Jose. You can search, favorite lines/stops, and see live countdowns with minimal friction. It’s built entirely in SwiftUI using native Apple frameworks. Happy to answer questions about the API, SwiftUI quirks, or anything else. Feedback welcome! https://ift.tt/hv0EYi8 April 19, 2025 at 12:30AM

Show HN: I made a Doom-like game fit inside a QR code https://ift.tt/60dCyFc

Show HN: I made a Doom-like game fit inside a QR code https://ift.tt/vjTzUoM April 18, 2025 at 11:40PM

Show HN: Attune - Build and publish APT repositories in seconds https://ift.tt/d1X8Dcq

Show HN: Attune - Build and publish APT repositories in seconds Hey HN, we're Eliza and Xin, and we’ve been working on Attune. Attune is a tool for publishing Linux packages. Previously, we worked at other startups building open source developer tools that ran on our customers’ CI and development machines. For many of them, being able to `apt-get install` our tools was a requirement. When we went to actually set up APT repositories, we were really surprised by the state of tooling around package publishing. The open source tools we found were old, slow, and difficult to figure out how to run in CI. The commercial tools we found were not much better. The cloud-hosted vendors required us to provide our signing keys to a cloud vendor (which was a non-starter), while the self-hosted vendors required us to operate our own specialized hosting servers. We just wanted something simple: sign locally, run quickly, be easy to use, and deploy to managed object storage. We couldn’t find it, so we built it. If you want to try it out, you can create a repository with three commands: attune repo create --uri https://ift.tt/O1b9kl5 attune repo pkg add --repo-id 123 package.deb attune repo sync --repo-id 123 You can get the tool at https://ift.tt/suRC4Y0 . There are a lot of rough edges right now since it's so new - sorry in advance, we're working on sanding those down. It’s fully open source under Apache 2. We’re also working with some early customers to build enterprise features like audit logging, RBAC, and HSM integrations, and we’re thinking about building a managed cloud hosting service as well. We’d love your feedback on whether this is useful for you, and what you’d like to see next. We’re well aware that publishing is a small piece of CI/CD, but we think a lot of the tooling in this area (publishing, artifact registries, package repositories) could really use some love. What do you think? Comment here, or email us at founders@attunehq.com. https://ift.tt/suRC4Y0 April 18, 2025 at 11:14PM

Show HN: I made a Chrome extension to see what's working on any TikTok profile https://ift.tt/ruWSezj

Show HN: I made a Chrome extension to see what's working on any TikTok profile I built this because I was tired of manually checking TikTok profiles to understand what kind of content performs well. The extension lets you: 1) Analyze any TikTok profile 2) See engagement rate, average views, and post frequency at a glance 3) Sort videos by likes, comments, or shares - without clicking into each one 4) Download videos and thumbnails without watermarks 5) Export data if you want to keep a copy or do more digging It’s mostly for creators, marketers, or anyone curious about what’s working on TikTok. Would love your feedback, ideas, or thoughts on how to improve it. https://ift.tt/g60jWr7 April 18, 2025 at 11:10PM

Show HN: Image2video.app – Transform Ghibli-style images into videos using AI https://ift.tt/Ft7OPeA

Show HN: Image2video.app – Transform Ghibli-style images into videos using AI Hi HN, I’ve developed a web app that lets you turn static Ghibli-style images into animated videos. It uses AI to create smooth, dynamic animations from your uploads. I will also add Ghibli-style images generation abilities in the app soon. I’d love to hear your feedback and see what you create with it :) Check it out at https://ift.tt/ENsqr6A . https://ift.tt/ENsqr6A April 17, 2025 at 10:32PM

Show HN: Milter in Rust to Add Headers https://ift.tt/CEkVf7W

Show HN: Milter in Rust to Add Headers Here's a milter in Rust that adds List-Unsubscribe headers. It creates a URL that has encoded email-from, rcpt-to and a HMAC SHA 256 verification hash using a shared secret key. Possibly it improves delivery of newsletters and transactional emails. https://ift.tt/aKrsCTh April 17, 2025 at 03:52AM

Show HN: LibreAI – Private AI chat app with open-source models (no ChatGPT) https://ift.tt/lzZ3nr5

Show HN: LibreAI – Private AI chat app with open-source models (no ChatGPT) I built LibreAI to provide a private, self-hostable alternative to ChatGPT using open-source models like Mistral, LLaMA 3, and Phi via Ollama. It's fast, streams responses in real-time (via HTMX), and has zero telemetry. There’s a free tier to try it out and a paid option to support development. Would love feedback, feature requests, or model suggestions! Built with Go Fiber + HTMX, hosted on my own infra. No JavaScript frontends here :) Live at: https://libreai.app https://libreai.app April 16, 2025 at 09:53PM

Show HN: Particle - News, Organized https://ift.tt/JoQab79

Show HN: Particle - News, Organized Hello HN! Particle News product engineer here. Keeping up with the news is overwhelming in an age of information overload. Particle reimagines the experience by organizing articles into comprehensive "Stories," offering clear, concise summaries to quickly grasp what matters. Today, we reached the #1 spot in "Newspapers & Magazine" on the iOS App Store—and I thought I'd share a bit of our backstory. I've been connected to this team for a long time. About 20 years ago, I shared a house with our CTO and co-founder Marcel Molina. I helped him get started with programming. Since then, Marcel has had an extraordinary career—becoming a senior staff engineer at Twitter, where he helped build foundational features like Retweets, Notifications, and Lists, and later working at Tesla on manufacturing execution systems that scaled across Gigafactories. At Twitter, Marcel worked closely with our CEO and product visionary Sara Beykpour, who led initiatives like Twitter Blue, Twitter Video, and the experimental app twttr . Sara has a background in Software Engineering and Cognitive Science from the University of Waterloo and spent over a decade at Twitter in engineering and leadership roles. In late 2022, Sara and Marcel started prototyping a news app that could reduce the cognitive and emotional burden of staying informed—by using AI to help people understand more, faster. They were soon joined by a few other former Twitter colleagues who helped shape the early concept into a working iOS application. I joined about 15 months ago to contribute across the entire stack. Since then, I've helped design and build major iOS features, rewritten our public website on Cloudflare Workers, and implemented new functionality in our Go backend, which is driven by Google Cloud's Pub/Sub architecture. What Makes Particle News Different Particle helps you navigate the news effortlessly—leveraging AI to help you understand more, faster. Some highlights: • Personalized News – Your feed is tailored to your interests. You can follow specific people, places, and things so you never miss what matters to you. • Clear Summaries – Get a quick overview or dive deeper with detailed, structured context—summarized in natural language. • Perspective Tools – Features like "Opposite Sides" and our political spectrum chart let you explore stories through multiple lenses. • Interactive Q&A – Ask questions about any story and get concise answers with sources and citations. • Audio Summaries – Use the "Play" feature to listen to your feed, specific stories, or even select articles—great for hands-free or on-the-go moments. One of the things we're most proud of is how Particle supports publishers. We've partnered with outlets like Reuters, AFP, and Fortune to host some of their content via APIs. These partners get prominent placement, and their links are highlighted in gold to stand out. This model aims to drive traffic back to publishers and reward high-quality journalism, rather than just aggregating and commodifying it. Transparency is a core value: all sources are cited, generated answers are grounded in evidence, and we take real care to prevent AI hallucinations or misleading summaries. Despite negligible marketing spend, Particle has grown to the top of its category by focusing on engagement with early users and meaningful partnerships with the media ecosystem. Coming soon: weekday mini crosswords—a new feature designed by another longtime friend of ours from 20 years back who went on to work at Twitter, lead development on Firewatch, and release his own games independently. It's incredibly fun and rewarding to be building something meaningful with old and new friends. I feel lucky every day to work alongside some of the best product, design, and engineering minds on a project we hope will help people stay engaged with democracy without burning out. https://ift.tt/lrso3Jw April 16, 2025 at 04:26AM

Show HN: Torque – A lightweight meta-assembler for any processor https://ift.tt/C3gahIJ

Show HN: Torque – A lightweight meta-assembler for any processor Hello everyone, I've been working on this project for the past few months. Torque is a meta-assembler: instead of having an instruction set built into the assembler, you use macros to build up a small language that decribes an instruction set and then you use that to write your program. It's designed to work for any microcontroller/processor architecture, you build from the bit level upwards so there aren't any assumptions around word widths, instruction formats, or endianness. I created Torque initially to write programs for a PIC microcontroller, after running into difficulties with the official assembler. I've also used it to write programs for the Z80 processor inside an old TRS-80 computer. Let me know if you try it out or have any questions! https://ift.tt/Xfj0a8B April 16, 2025 at 04:46AM

Show HN: I asked Gemini2.5 pro to create a "bro version" of a stoic book https://ift.tt/OqryjLU

Show HN: I asked Gemini2.5 pro to create a "bro version" of a stoic book https://ift.tt/x5UQTpR April 16, 2025 at 12:36AM

Show HN: Chonkie Cloud – No-nonsense chunking now on the the cloud https://ift.tt/rj0x1Qe

Show HN: Chonkie Cloud – No-nonsense chunking now on the the cloud Hey HN, We’re the developers of Chonkie, a powerful and easy-to-use chunking library. Last November, we introduced Chonkie OSS here on HN ( https://ift.tt/ISwQGsk ) and got some fantastic feedback. Today, we’re excited to share Chonkie Cloud, a fully hosted service for chunking, along with new tools to make building and debugging RAG systems easier than ever. --- A quick recap for context: We launched Chonkie as an open-source project late last year. A few weeks ago, we decided to go full-time on it. Unfortunately, this shift wasn’t as smooth as we had hoped. Due to some legal stuff, we had to rebuild the entire project from scratch in a new repo. Restarting sucked but it gave us the chance to clean things up and build something faster, cleaner, and better. You can check out the new repo here: https://ift.tt/AEgcv49 --- What we're launching today: - Chonkie Cloud – A fully hosted chunking service. Just send in your texts and get chunks back. It supports all our chunkers and comes with a playground to help visualize RAG. Try it out: https://chonkie.cloud (no credit card required) - Chonkie Recipes – Predefined chunking configs to help you quickly get started. If you're not sure which settings to use, just find the closest match in our recipes. These work with the Recursive and Late chunkers. Available here: https://ift.tt/lMZWwyT - Chonkie Vizard – A tool to visualize chunking output during development. If you've ever debugged with print statements full of dashes and stars, this is for you. Installation: pip install "chonkie[viz]" Example usage (1) : viz.print(chunks) # Print in terminal Example usage (2): viz.save("out.html", chunks) # Save for later --- What's next: - A code chunker - File “chefs” to help clean and prepare data - Integrations with vector DBs and data extraction tools - Support for more languages (typescript is #1 on our list) --- We’d love feedback on what we’re building. If you have feature requests or ideas, please open an issue on the repo ( https://ift.tt/AEgcv49 ) or comment here. Thank you! https://ift.tt/MVzTHKD April 15, 2025 at 11:06PM

Show HN: Kforward – Lightweight K8s proxy for local dev https://ift.tt/i8vW5Zb

Show HN: Kforward – Lightweight K8s proxy for local dev https://ift.tt/AB5oakN April 14, 2025 at 09:44PM

Show HN: Go library for generating Anki decks https://ift.tt/tfXmxEZ

Show HN: Go library for generating Anki decks https://ift.tt/zBxZlIG April 14, 2025 at 09:33PM

Show HN: LeetGPU, a playground to learn practice and hone your GPU programming. https://ift.tt/9UYjTP2

Show HN: LeetGPU, a playground to learn practice and hone your GPU programming. Learn, write, practice CUDA programming on LeetGPU.com, an online CUDA playground for anyone to write and execute CUDA code without needing a GPU and for free April 13, 2025 at 11:02PM

Show HN: Gatehouse-TS – TypeScript port of Rust's authorization policy framework https://ift.tt/W1Su0dI

Show HN: Gatehouse-TS – TypeScript port of Rust's authorization policy framework TL;DR Gatehouse-TS is a zero-dependencies authorization policy library written in TypeScript, ported from Rust. It allows you to combine different access control policies to check before allowing someone to access a resource. A few weeks ago, I saw a post on HN about Gatehouse, an authorization library written in Rust. Here's the original HN thread: https://ift.tt/gajnCLu I looked at their GitHub page, and started reading the source code. I do this for anything posted here that catches my attention, even if it's written in a language I don't understand. The code was relatively easy to follow, even though I'm not a Rust programmer. Why did I decide to try rewriting it in TypeScript? For fun, of course! That was basically the only reason. I had nothing to do on a weekend and wanted to write code. It _did_ help with some challenges I was facing at $day_job, but that came afterwards. I emailed the authors of the Rust version, telling them that I was working in a TypeScript version, and they kindly blessed me to publish it with a similar name. Overall I'm very happy with it. The original authors provided an amazing foundation for a port, I had fun rewriting it and trying to adapt their APIs from Rust to TypeScript. As a bonus, it was immediately useful at work as well. I appreciate any feedback or thoughts on it. Please check out the original project as well, since most of the hard work was done by them: https://ift.tt/kIxyePN Documentation: https://ift.tt/yAFTVUR https://ift.tt/32mgVjQ April 13, 2025 at 11:39PM

Show HN: Opsmate - LLM Powered SRE Assistant https://ift.tt/e9vdFXs

Show HN: Opsmate - LLM Powered SRE Assistant https://ift.tt/JOv0Ppl April 13, 2025 at 10:49PM

Show HN: AI quiz generator from any topic or book in seconds https://ift.tt/aACbKyU

Show HN: AI quiz generator from any topic or book in seconds https://www.wiyomi.com April 10, 2025 at 12:27PM

Show HN: Downr – An All-in-One Social Media Downloader for 50 Platforms https://ift.tt/auW3kwc

Show HN: Downr – An All-in-One Social Media Downloader for 50 Platforms https://downr.org April 12, 2025 at 06:06PM

Show HN: OctAPI – Visualize API Routes Directly in VS Code https://ift.tt/5BnW9xF

Show HN: OctAPI – Visualize API Routes Directly in VS Code Started noticing issues while working with friends that I thought only I was encountering: spending too long manually adding routes to postman, getting lost looking for a single route in the sea of files and thought Hey, I can fix that (I think?) So OctAPI is a VSCode extension that automatically detects and displays API routes from a local project (currently supports Express.js, NestJS, Next.js, Koa, Flask, and FastAPI), it filters and groups them, generates a Postman export JSON file, and adds to favs. It doesn't run your code and it all stays in your machine A friend literally screamed when they were live testing it for me lol It's open source ( https://ift.tt/oYrMSyz ) and here's its landing page ( https://ift.tt/WHMfC0z ) I'm hoping to add custom tags, more frameworks support and even AI features. https://ift.tt/Tr7zMOd April 13, 2025 at 03:18AM

Show HN: memEx, a personal knowledge base inspired by zettlekasten and org-mode https://ift.tt/dI9q4P0

Show HN: memEx, a personal knowledge base inspired by zettlekasten and org-mode https://ift.tt/qjTFGfc April 13, 2025 at 02:02AM

Show HN: Atari Missile Command Game Built Using AI Gemini 2.5 Pro https://ift.tt/HyeFkib

Show HN: Atari Missile Command Game Built Using AI Gemini 2.5 Pro A modern HTML5 canvas remake of the classic Atari game from 1980. Defend your cities and missile bases from incoming enemy attacks using your missile launchers. Initially built using the Google Gemini 2.5 Pro AI LLM model. https://ift.tt/vZ73hzq April 7, 2025 at 01:18PM

Show HN: Lunon – Instant model switching across LLMs https://ift.tt/bUrPA9n

Show HN: Lunon – Instant model switching across LLMs Hey HN! We built Lunon to make LLM development way less of a headache. Ever wanted to see how different models handle the same prompt without all the setup hassle? That's what we fixed. Our API lets you compare Claude, GPT, Mistral and others in real-time with just a few lines of code. No more complex infrastructure or managing multiple API connections - we handle all that boring stuff behind the scenes. Plus, you can cut costs by intelligently routing requests to the right model for each task. Use the powerful (expensive) models only when you really need them. If you're building with LLMs and tired of the integration headaches, would love to hear feedback! https://lunon.com/ April 12, 2025 at 01:16AM

Show HN: A gallery of ChatGPT image styles for inspiration https://ift.tt/cMFgZPs

Show HN: A gallery of ChatGPT image styles for inspiration I found myself collecting a ton of different styles to explore and get inspiration for other photos so I built this to share the collection with others. The site was also completely coded using ChatGPT. https://imagestyles.ai April 12, 2025 at 12:40AM

Show HN: The Mouse is Lava~ A Chrome Ext to navigate websites with your keyboard https://ift.tt/yJSUsFX

Show HN: The Mouse is Lava~ A Chrome Ext to navigate websites with your keyboard https://ift.tt/ODcIVCF April 11, 2025 at 11:09PM

Show HN: I built a tool to manage and compare credit card rewards https://ift.tt/3s5v6GY

Show HN: I built a tool to manage and compare credit card rewards This is a free tool that helps you manage and visualize your credit card rewards across different categories. You can input the cards in your wallet and see how they complement each other, spot gaps in your setup, and also see the best card to use for a given merchant. I’m also a founder at OneCard, where we’re building a smart card that’ll eventually handle all of this automatically, routing each purchase to the best card in real-time. Would love feedback from the HN community! https://ift.tt/JBYevqc April 11, 2025 at 04:24AM

Show HN: Calculate confidence score for OpenAI JSON output https://ift.tt/KmAV5Cd

Show HN: Calculate confidence score for OpenAI JSON output https://ift.tt/fotU7Gs April 10, 2025 at 10:16PM

Show HN: I built an app that reduces podcast preparation effort by 95% + https://ift.tt/XkZcqUD

Show HN: I built an app that reduces podcast preparation effort by 95% + https://ift.tt/ul2qbPc April 11, 2025 at 12:52AM

Show HN: Aqua Voice 2 – Fast Voice Input for Mac and Windows https://ift.tt/kU4ulrm

Show HN: Aqua Voice 2 – Fast Voice Input for Mac and Windows Hey HN - It’s Finn and Jack from Aqua Voice ( https://withaqua.com ). Aqua is fast AI dictation for your desktop and our attempt to make voice a first-class input method. Video: https://ift.tt/gxLuZEY Try it here: https://ift.tt/8ac9tRG Finn is uber dyslexic and has been using dictation software since sixth grade. For over a decade, he’s been chasing a dream that never quite worked — using your voice instead of a keyboard. Our last post ( https://ift.tt/BYPub3g ) about this seemed to resonate with the community - though it turned out that version of Aqua was a better demo than product. But it gave us (and others) a lot of good ideas about what should come next. Since then, we’ve remade Aqua from scratch for speed and usability. It now lives on your desktop, and it lets you talk into any text field -- Cursor, Gmail, Slack, even your terminal. It starts up in under 50ms, inserts text in about a second (sometimes as fast as 450ms), and has state-of-the-art accuracy. It does a lot more, but that’s the core. We’d love your feedback — and if you’ve got ideas for what voice should do next, let’s hear them! https://withaqua.com April 9, 2025 at 11:31PM

Show HN: I built a tool that deconstructs websites to reveal their tech stack https://ift.tt/noKO2xy

Show HN: I built a tool that deconstructs websites to reveal their tech stack Hi HN, I built https://unbuilt.app to solve a problem I frequently faced as a developer: identifying the technology stack behind websites, especially those using newer frameworks and tools. While existing solutions rely on pre-saved data or signature databases, Unbuilt performs a fresh analysis by actually loading and examining the website code in real-time. This means it can detect cutting-edge technologies that often get missed by other analyzers. Technical details: - Uses Playwright to visit sites and analyze their resources - Queue-based architecture for handling concurrent requests - Optimized patterns for detecting modern frameworks (Next.js, Vite, React Compiler, etc.) - Dual-layer caching system for both performance and result sharing The tool is completely open-source and free to use. There are no premium tiers or usage limits. I built it because I believe developers deserve better tools for understanding the web ecosystem. I'd appreciate any feedback, particularly on detection accuracy and the types of technologies you'd like to see added. If you're interested in contributing, all pattern detection logic is designed to be easily extensible. Link to repo: https://ift.tt/2iA8uwc https://unbuilt.app/ April 10, 2025 at 12:27AM

Show HN: eInk Manga with Kindle Comic Converter (+Kobo/ReMarkable) https://ift.tt/g4IKbBX

Show HN: eInk Manga with Kindle Comic Converter (+Kobo/ReMarkable) https://ift.tt/2xeRV6z April 9, 2025 at 11:26PM

Show HN: Coroot – eBPF-based, open source observability with actionable insights https://ift.tt/y9quaDp

Show HN: Coroot – eBPF-based, open source observability with actionable insights A common open source approach to observability will begin with databases and visualizations for telemetry - Grafana, Prometheus, Jaeger. But observability doesn’t begin and end here: these tools require configuration, dashboard customization, and may not actually pinpoint the data you need to mitigate system risks. Coroot was designed to solve the problem of manual, time-consuming observability analysis: it handles the full observability journey — from collecting telemetry to turning it into actionable insights. We also strongly believe that simple observability should be an innovation everyone can benefit from: which is why our software is open source. Features: - Cost monitoring to track and minimise your cloud expenses (AWS, GCP, Azure.) - SLO tracking with alerts to detect anomalies and compare them to your system’s baseline behaviour. - 1-click application profiling: see the exact line of code that caused an anomaly. - Mapped timeframes (stop digging through Grafana to find when the incident occurred.) - eBPF automatically gathers logs, metrics, traces, and profiles for you. - Service map to grasp a complete at-a-glance picture of your system. - Automatic discovery and monitoring of every application deployment in your kubernetes cluster. We welcome any feedback and hope the tool can improve your workflow! https://ift.tt/Dvb8JPo April 8, 2025 at 11:49PM

Show HN: HNSW index for vector embeddings in approx 500 LOC https://ift.tt/sIAfOXH

Show HN: HNSW index for vector embeddings in approx 500 LOC https://ift.tt/NI6apTP April 8, 2025 at 10:43PM

Show HN: Zev – Remember (or discover) terminal commands https://ift.tt/amhxKFG

Show HN: Zev – Remember (or discover) terminal commands I'm often in the terminal and forget a command (or never knew it in the first place). This is a simple tool that lets you type out what you want, gives you several options, and then automatically puts the one you want in your clipboard. https://ift.tt/YDsSAZx April 8, 2025 at 10:26PM

Show HN: Daily Vim – Master Vim with Spaced Repetition https://ift.tt/0GyUor7

Show HN: Daily Vim – Master Vim with Spaced Repetition Hey HN, After years of using Vim, I still found myself forgetting useful commands. There are so many hidden tricks, and mastering them all at once is overwhelming… I’d learn a great trick, only to realize weeks later that I had completely forgotten it. I tried everything—cheatsheets, tutorials, even keeping a notebook. But nothing really stuck. So I built Daily Vim to solve this problem: - Each day, you get a bite-sized, actionable Vim tip. - Rate your familiarity, so the system adapts to you. - Tips follow spaced repetition, helping you remember them long-term. Why? Vim is powerful, but hard to master. Most resources dump everything on you at once. Daily Vim is built to reinforce what you learn, so you can stop forgetting useful commands. Who’s it for? - Beginners struggling to get past `:q!` - Experienced users who want to expand their toolkit - Power users who don’t want to forget what they’ve learned I’d love to hear your thoughts! What’s your biggest struggle with Vim? https://dailyvim.com April 8, 2025 at 07:55PM

Show HN: SQLTest.online: A free, interactive SQL practice platform I built https://ift.tt/IDl5dv8

Show HN: SQLTest.online: A free, interactive SQL practice platform I built Just launched: SQLTest.online - a free, hands-on platform for learning and practicing SQL. Features exercises covering basic syntax through advanced topics. This is a personal project I've been working on. https://sqltest.online April 7, 2025 at 11:35PM

Show HN: Lately ADHD Journey Planner https://ift.tt/xy6sdAb

Show HN: Lately ADHD Journey Planner https://ift.tt/JiZ9S8C April 7, 2025 at 10:32PM

Show HN: Browser MCP – Automate your browser using Cursor, Claude, VS Code https://ift.tt/XRAy7NY

Show HN: Browser MCP – Automate your browser using Cursor, Claude, VS Code https://browsermcp.io/ April 7, 2025 at 11:25PM

Show HN: Copy Folder to ChatGPT in 0.6s – OpenRepoPrompt https://ift.tt/2QZxqGw

Show HN: Copy Folder to ChatGPT in 0.6s – OpenRepoPrompt Over this past month I had the idea to build a 100% open-source MIT-licensed tool to simplify sharing code with LLMs, without the vendor lock-in you get from most SDKs. Right now, it’s way too hard to export your data or work freely with models like o1 PRO or Grok 3, especially since they don’t even have API access. So I built OpenRepoPrompt, an open-source tool from wildberry-source that serializes files and folders into XML for LLMs. I coded/designed from 12PM -> 11PM on Saturday and 8AM -> 11PM on Sunday. There are still tons of features missing (I'm working on better file filtering next!) but I think it's reached that - buzzword alert! - MVP stage where I want to see if anyone's actually going to use the darn thing. Here's the OpenPrompt repo: https://ift.tt/5kNnuG1 . Here are some screenshots I took while making it: https://ift.tt/YyNg5TH... . I think it has some potential. What do you guys think? https://ift.tt/5kNnuG1 April 4, 2025 at 06:14PM

Show HN: A fast, minimal and offline-friendly web playground https://ift.tt/SFBv5UM

Show HN: A fast, minimal and offline-friendly web playground I built a web-based HTML/CSS/JS editor focused on speed, simplicity, and offline access. No bloat — just open and start coding. What makes it different: Live Preview – Edit HTML, CSS, and JS side-by-side with instant feedback. Offline support – Works without internet. You can even install it as a PWA and use it like a native app. No Login Required – Just visit, code, and preview. Login only if you want to save/share. Savable & Shareable Links – Save scripts in the cloud and get shareable links. Customizable Editor – Themes, fonts, auto-format on save, layout tweaks, line wrapping, etc. Hotkey Support – Power user shortcuts with tooltips showing keybinds. Download as ZIP – Export your project with HTML/CSS/JS files separated. Ideal for tinkering, prototyping, teaching, or even building micro-tools. It’s intentionally simple and fast — more features coming soon. Try a sample snippet: https://ift.tt/xoPSzpa https://ift.tt/yuOa5ne April 7, 2025 at 03:25AM

Show HN: Latitude.sh Databases – Simple PostgreSQL DBaaS on Bare Metal https://ift.tt/TQgkyLw

Show HN: Latitude.sh Databases – Simple PostgreSQL DBaaS on Bare Metal Hi HN, Gabriel here, one of the developers at Latitude.sh (we're a bare metal cloud provider). Over the past year or so, I've been the primary developer building Latitude.sh Databases – our take on a managed PostgreSQL service. The core idea was to offer a straightforward and competitively priced option for developers who need reliable PostgreSQL without overly complex configurations, leveraging the performance benefits of running directly on bare metal. It runs on our global bare metal infrastructure. Key features we've implemented include: * Built-in monitoring & connection pooling * IP Address Whitelisting (Trusted Sources) * Automated backups configured directly to your own S3 bucket (giving you control over storage and potentially costs) * An optional integration with Supabase, allowing you to use parts of their dashboard for enhanced usability with your database. Under the hood, it's built on Kubernetes running on our bare metal servers, using the CloudNativePG operator to manage the PostgreSQL instances. We've found this operator approach works well for handling stateful database workloads in K8s, challenging the old notion that databases don't belong there. The service stemmed from internal needs and early interest gathered from a waitlist (~300 signups). It's currently live and available for use. We're launching it here on HN because we'd genuinely appreciate your feedback on: * The overall developer experience and UI simplicity. * The current feature set (especially the S3 backup and Supabase integration). * Performance perceptions (given the bare metal base). * Our pricing model's competitiveness and clarity. * Any technical aspects of the implementation (running PG on K8s/bare metal). Happy to answer any questions you have! You can check out the product page here: https://ift.tt/jraO5Sc Thanks for looking! https://ift.tt/jraO5Sc April 7, 2025 at 01:00AM

Show HN: I have built my first AI notetaker. Any feedback? https://ift.tt/w7hcNy2

Show HN: I have built my first AI notetaker. Any feedback? Hey there, While searching for my next role as a PM in Blockchain + AI, I started building an AI-powered notetaker called MindNote: https://ift.tt/UhsWx1w . The idea is simple: you input your thoughts, and AI helps you complete or modify them. If you decide to test it, try using the prompt: Keep this text and…"(add budget, complete list, organize alphabetically, add their translation, etc.). The first big release is all about modifying your notes with AI prompts. The next big step? Capturing thoughts directly from the mind using brain-to-text technology. There is a 3 days free trial: https://ift.tt/UhsWx1w Video tutorial: https://www.youtube.com/watch?v=okq_LZLbDfI April 6, 2025 at 11:06PM

Show HN: I made an AI Platform that gamifies applying to jobs https://ift.tt/7sRMdiL

Show HN: I made an AI Platform that gamifies applying to jobs Hi there! I've created ApplyNinja, a platform that uses AI to help jobseekers apply to jobs faster and better. What differentiates ApplyNinja from the other "auto-apply" platforms is the trained AI model behind it. It learns from your resume and the jobs you apply to, and it provides better and better results. Currently, it serves as a helper for the entire application process: - Track & Monitor job applications - Generates Resume suggestions and highlights the parts that are not so good - Auto-fixes and generates a new resume based on the suggestions - Creates Cover Letters based on your resume and the job description - Generates Company Insights, scraping user reviews from Glassdoor, Indeed etc. - Generates Salary Insights so jobseekers have a negotiation leverage - Generates Technical Interview Questions based on the Job Description On top of all of this, it provides a gamified experience, because consistency is the key in applying to jobs. Me and my friends are using it daily and manage to land approximately 5-10 interviews/month as Software Engineers. I would really love any kind of feedback. Thank you very much! https://ift.tt/RoYCB3j April 6, 2025 at 01:39AM

Show HN: Experience iPhone's "Blob Keyboard" prototype from 2005 https://ift.tt/W1jCTH5

Show HN: Experience iPhone's "Blob Keyboard" prototype from 2005 Hi HN, I teach tech design history, and one of the key stories I cover is the development of the original iPhone keyboard by Ken Kocienda. Reading about it in his book "Creative Selection" is great, but I wanted my students (and now you!) to actually feel this step in the process. So, I built a web simulator of the "Blob Keyboard", Kocienda's very first attempt at a touchscreen keyboard that actually works, from September 2005: Try the Blob Keyboard: https://ift.tt/zkdXKt1... - Tap for the middle letter - Swipe left or right for the side letters More on the github repo: https://ift.tt/vmVuld2 The Blob Keyboard prototype emerged during a UX crisis for iPhone team (their software keyboard just didn't work at all, fingers being too big, and the Newton failure loomed over them), highlighting how innovation is rarely a straight path. It was developed on a tethered touchscreen display codenamed "Wallaby". To make this simulator as authentic as possible, I referenced images from Kocienda's book and even got direct feedback and guidance from Ken Kocienda himself on Bluesky. What to expect (or… what not to expect): This is a reconstruction of a very early prototype with limitations reflecting that specific moment. The goal was to test first if typing with accuracy was even possible, as all the rest was moot if it failed! It's NOT QWERTY: They were still hoping to get us out of QWERTY, but then familiarity won. No Backspace: You can't delete. No Cursor Movement: The text field is just a simple display. No Caps or Numbers: Only lowercase letters. No Smooth Animations: Keys just "pop" instantly when pressed. Kocienda noted that your eye fills in the gaps, giving a sense of movement. Best Experience: While it works with a mouse/trackpad on desktop, it's designed for touchscreens to better replicate the original Wallaby hardware interaction. Use it on your phone! This project aims to provide a tangible glimpse into a turning point moment in iPhone development and the iterative nature of design. It's like stepping back in time and trying out that early demo on Kocienda's desk. I would love to hear your reactions and thoughts on experiencing this piece of UI history! What other significant prototype do you wish you could experience? April 6, 2025 at 01:20AM

Show HN: Reharmonizing Racing in the Street by Bruce Springsteen https://ift.tt/in0NOUc

Show HN: Reharmonizing Racing in the Street by Bruce Springsteen I put together a short essay on re-harmonizing Racing in the Street by Bruce Springsteen. Just a simple demonstration of what’s possible today with a few hours and the right mix of technology https://ift.tt/WBxLS2i April 6, 2025 at 12:41AM

Show HN: Teletyped – Free interface for multiple AI models https://ift.tt/2lQypgC

Show HN: Teletyped – Free interface for multiple AI models We built a clean UI for ChatGPT and Claude that's now completely free. Just plug in your API keys and get: - Switch models mid-chat - Edit AI responses - Saved & temporary chats - Adjustable temperature - Multiple themes No subscriptions. We made this because we were tired of switching between different AI interfaces. Would love your feedback! https://ift.tt/3fh2UYs April 5, 2025 at 09:23PM

Show HN: DuckDB Powered SQL Editor https://ift.tt/Ih7Dz41

Show HN: DuckDB Powered SQL Editor Hi guys, I made an SQL editor that utilizes the duckDB engine to process your queries. As a result, the speed gains are +25% when compared to using any standard editor that connects through JDBC. I built this because I work on a small data team and we can't justify an OLAP database. Postgres is amazing but, if I try to run any extremely complex queries I get stuck waiting for several minutes to see the result. This makes it hard to iterate and get through any sort of analysis. That's when I got the idea to use duckDB's processing engine rather than the small compute available on my Postgres instance. I didn't enjoy writing SQL in a Python notebook and wanted something like dBeaver that just worked, so I created soarSQL. Try it out and let me know if it has a place in your toolkit! https://soarsql.com April 5, 2025 at 03:10AM

Show HN: I made an online free tool site https://ift.tt/r5xysk3

Show HN: I made an online free tool site Hi HN! I've created [UFreeTools]( https://ift.tt/HCsi3Sv ), a collection of developer utilities (60+ tools and growing) that's completely free, ad-free, and runs entirely in the browser. No data leaves your device - everything is processed locally. ## Background As a developer, I found myself constantly searching for simple utilities like JSON formatters, UUID generators, and color pickers. Most existing options were cluttered with ads, required sign-ups, or sent data to servers. I wanted clean, fast tools that respected privacy. So I built UFreeTools as a modern SPA using Vue 3 and Vite with: - 100% client-side processing for all tools - Progressive loading with dynamic imports (only load what you use) - Dark/light theme support - Offline capability for most tools - Responsive design that works well on mobile ## Technical Details Some of the more technically interesting implementations: - *JWT Debugger*: Uses WebCrypto API for verification of all common algorithms - *Symmetric Encryption*: Implements AES-GCM/CBC/CTR with PBKDF2 key derivation - *Image Processing*: Uses Web Workers for non-blocking operations on large images ## Lessons Learned The biggest challenges were: 1. *Performance optimization*: Some tools (like image processors) needed careful optimization to handle large files without freezing the UI 2. *Browser API limitations*: Working around browser restrictions for certain crypto operations 3. *Bundle size management*: Keeping the initial load small while supporting 60+ tools ## Future Plans I'm planning to add: - Tool configurations sync via local storage - More specialized tools for developers based on feedback I built it because I needed these tools myself, and I hope others find it useful. I'd love your feedback, especially on UX, tool suggestions, or if you find any bugs! [Try UFreeTools]( https://ift.tt/HCsi3Sv ) https://ift.tt/HCsi3Sv April 5, 2025 at 12:56AM

Show HN: I made a platform to load and performance test API's https://ift.tt/PhNvomu

Show HN: I made a platform to load and performance test API's I built BuzzBench because I was frustrated with how complex performance testing tools have become. And I was ending up writing my own scripts to test endpoints and manually check out resource usage at the time of testing. Checkout demo: https://www.youtube.com/watch?v=yAnbZMoQvmQ https://buzzbench.io/ April 2, 2025 at 05:52PM

Show HN: Portowl – Simple Retro RPG https://ift.tt/P89XkDc

Show HN: Portowl – Simple Retro RPG Experience first-hand the wonderful world of owl photography! Owls aren’t big on having their photo taken, so learning how may take some time. While you play, be on the lookout for powerful upgrades. https://ift.tt/G0kRWuA April 4, 2025 at 11:52PM

Show HN: Haystack Code Reviewer – Guided Tours for Pull Requests https://ift.tt/ABLQvJO

Show HN: Haystack Code Reviewer – Guided Tours for Pull Requests Hi HN! We’re excited to share Haystack Code Reviewer, a tool that analyzes, chunks, and organizes the changes made in a pull request and guides you through them in a logical sequence on an interactive canvas. We hope that Haystack really cuts down on the time that you spend piecing together what's happening in the pull request, so that you can spend time on the parts that matter, like providing feedback or internalizing the changes. As software engineers who’ve dreaded code reviews and always put them off as much as possible, we noticed how much time we spent poring over diffs to understand the broader picture or narrative of a pull request before even evaluating the code itself. Haystack Code Reviewer is designed to assemble that jigsaw puzzle for you, making code reviews less of a chore and more of a learning opportunity for both the reviewer and author! To play with demo code reviews, go to https://ift.tt/murBkb8 and select a pull request of your choice. If you’re interested in learning more, feel free to check out our short demo video at https://youtu.be/K_qLwXFwr8I . Or if you’re ready to try it out, head over to https://ift.tt/Jvq9Mfu and: 1. Optionally sign in (necessary if you want to review a pull request for a private repository). For public repositories, no sign-in is needed 2. Load any PR via URL or from the dropdown (for example, try https://ift.tt/vAVNZ10 ) 3. When prompted for AI analysis, click “Yes” to enable automatic organization and sectioning (we send only the diffs in the pull request to OpenAI). We hope you find Haystack Code Reviewer useful. Let us know what you think! You might have seen our previous Show HN at https://ift.tt/uvNcO2Z . We have now completely redesigned it to work with a guided AI tour of exactly how you should review the pull request so that you don't have to even think about how to approach a code review! https://ift.tt/murBkb8 April 3, 2025 at 11:16PM

Show HN: Synto - micro-learning app to help you deeply understand classic books https://ift.tt/lWCdytf

Show HN: Synto - micro-learning app to help you deeply understand classic books https://ift.tt/x5SNEVd April 3, 2025 at 11:14PM

Show HN: Arrakis – Open-source, self-hostable sandboxing service for AI Agents https://ift.tt/DRAVM5p

Show HN: Arrakis – Open-source, self-hostable sandboxing service for AI Agents Hi HN! My name is Abhishek. I've spent my career working on Operating Systems and Infrastructure at places like Replit, Google, and Microsoft. Today, I'm excited to launch Arrakis: an open-source and self-hostable sandboxing service designed to let AI Agents execute code and operate a GUI securely. GitHub: https://ift.tt/PQOp8dV Watch Claude build a live Google Docs clone using Arrakis via MCP – with no re-prompting or interruption. https://www.youtube.com/watch?v=IZ5cAnhAdPQ Key Features - Self-hostable: Run it on your own infra or Linux server. - Secure by Design: Uses MicroVMs for strong isolation between sandbox instances. - Snapshotting & Backtracking: First-class support allows AI agents to snapshot a running sandbox (including GUI state!) and revert if something goes wrong. - Ready to Integrate: Comes with a Python SDK ( https://ift.tt/FfMYXOe ) and an MCP server ( https://ift.tt/l8h0JDy ) out of the box. - Customizable: Docker-based tooling makes it easy to tailor sandboxes to your needs. Sandboxes = Smarter Agents As the demo shows, AI agents become incredibly capable when given access to a full Linux VM environment. They can debug problems independently and produce working results with minimal human intervention. I'm the solo founder and developer behind Arrakis. I'd love to hear your thoughts, answer any questions, or discuss how you might use this in your projects! Get in touch - Email: abshkbh AT gmail DOT com - LinkedIn: https://ift.tt/9JofbCP Looking forward to your feedback! https://ift.tt/PQOp8dV April 3, 2025 at 12:08AM

Show HN: Agent File (.af) – An open file format for agents https://ift.tt/ocafARt

Show HN: Agent File (.af) – An open file format for agents Hi HN - We’re building Agent File (.af), which makes it possible to re-create the exact same agent (including memories, tools, message history, configs, etc.) across different machines. A big difference between LLMs and agents is that agents have associated state: system prompts, editable memory (personality and user information), tool configurations (code and schemas), and LLM/embedding model settings. While you can run the same LLM as someone else by downloading the weights, there’s no “representation” of agents that allows you to re-create an instance of an agent across services. Agent File (.af) is an open standard file format for serializing stateful agents. Originally designed for the Letta framework, .af is a human-readable representation of all the associated state of an agent to reproduce the exact behavior and memories. To demonstrate .af, we also made a few example agents with download links to .af: - MemGPT: An agent with memory management tools for infinite context, as described in the MemGPT paper Deep Research: A research agent with planning, search, and memory tools to enable writing deep research reports from iterative research - Customer Support: A customer support agent that has dummy tools for handling order cancellations, looking up order status, and also memory - Stateless Workflow: A stateless graph workflow agent (no memory and deterministic tool calling) that evaluates recruiting candidates and drafts emails - Composio Tools: An example of an agent that uses a Composio tool to star a GitHub repository We’d love to hear what people think of the agent schema we chose and if we’re missing anything (we included everything that we need from Letta, but there may be other features in other frameworks). https://ift.tt/6oWq2u0 April 2, 2025 at 11:44PM

Show HN: Mermaid Chart VS Code Plugin: Mermaid.js Diagrams in Visual Studio Code https://ift.tt/ljYMTGH

Show HN: Mermaid Chart VS Code Plugin: Mermaid.js Diagrams in Visual Studio Code https://ift.tt/teLMnyk April 2, 2025 at 11:33PM

Show HN: A Chrome extension to give you back control over short-form videos https://ift.tt/0fu8yWv

Show HN: A Chrome extension to give you back control over short-form videos Hi HN! I built this little extension to prevent, in my opinion, the most offensive anti-pattern used by tech companies. That is removing the seek bar in short-form videos. The "seek bar" is the bar at the bottom of a video that progresses as you play the video, and that you can click on or drag to skip around. Why companies ever thought it was a good idea to get rid of this I don't know, but I find it infuriating, so I decided to add it back for myself and thought others might like it too. ReelControl adds a progress bar and seeking capabilities to videos on Instagram, YouTube Shorts, and Facebook Reels. I do sometimes enjoy watching short-form content and I've found that with this extension enabled I can be more mindful about it and get sucked in way less. I'm also on my phone less because I tend to favor the web versions of these platforms now. Open source--PRs and issues welcome! https://ift.tt/CFDd8c9 https://ift.tt/7BDTOX3 April 1, 2025 at 06:51PM

Show HN: Monkeys.zip (Infinite Monkeys Writing Shakespeare) https://ift.tt/XQtUFlJ

Show HN: Monkeys.zip (Infinite Monkeys Writing Shakespeare) Hey HN! I've been working on this concept on and off for about a year, and challenged myself to actually release it in time for April Fools as I tend to just hoard projects. It's a simulator of the Infinite Monkey Theorem! You can browse our grid of monkeys that are busily typing away randomly on their keyboards, and see what words they've written - how many of those words appear in the works of Shakespeare. If you want, you can join in for free and get a monkey of your own. Monkeys get special rewards for typing special words - which can be used to customize their appearance. I just wiped the simulation this morning so we're starting from scratch! It's a little rough around the edges in some spots, if some content doesn't load right, just give it a refresh. I hope you get a kick out of it - it's been an absolute labor of love for me! https://monkeys.zip/ April 2, 2025 at 12:17AM

Show HN: I got burned by bad NYC rentals, so I built an extension for complaints https://ift.tt/DJxRTpn

Show HN: I got burned by bad NYC rentals, so I built an extension for complaints I built this after too many bad NYC apartment experiences: leaks, mice, no heat. All things that of course never show up on listings. Realest is a browser extension that overlays 311 housing complaint data on StreetEasy listings, right in the search results. It pulls from NYC’s open data and is meant to help renters avoid hidden issues. Would love feedback! https://realest.casa April 2, 2025 at 12:10AM

Show HN: Qwen-2.5-32B is now the best open source OCR model https://ift.tt/8Nr6msn

Show HN: Qwen-2.5-32B is now the best open source OCR model Last week was big for open source LLMs. We got: - Qwen 2.5 VL (72b and 32b) - Gemma-3 (27b) - DeepSeek-v3-0324 And a couple weeks ago we got the new mistral-ocr model. We updated our OCR benchmark to include the new models. We evaluated 1,000 documents for JSON extraction accuracy. Major takeaways: - Qwen 2.5 VL (72b and 32b) are by far the most impressive. Both landed right around 75% accuracy (equivalent to GPT-4o’s performance). Qwen 72b was only 0.4% above 32b. Within the margin of error. - Both Qwen models passed mistral-ocr (72.2%), which is specifically trained for OCR. - Gemma-3 (27B) only scored 42.9%. Particularly surprising given that it's architecture is based on Gemini 2.0 which still tops the accuracy chart. The data set and benchmark runner is fully open source. You can check out the code and reproduction steps here: - https://ift.tt/F7uc3le... - https://ift.tt/v3eDZw1 - https://ift.tt/ZCLfXcK https://ift.tt/MQh9t4u April 2, 2025 at 12:00AM

Show HN: Generate Anki Decks from YouTube Videos to Anki Converter https://ift.tt/iupG7tZ

Show HN: Generate Anki Decks from YouTube Videos to Anki Converter Hello! I love anki and have been using it for years to learn mandarin, but I'm lazy and always struggled to get over the "motivation hump" to sit down and create a deck. Recently I've been consuming a lot of AI content from youtube, and spent 3 days creating a deck for Andrej Karpathy's "Deep Dive into LLMs" video. This is impractical - so I built a little youtube to anki converter @ asimpleai.com It's rough around the edges... but I would like it to eventually become something useful to community. Feedback and suggestions are welcome! peace https://asimpleai.com/ April 1, 2025 at 11:17PM