2025年9月12日 の Show HN

After two years of development, I'm super excited to release Beat Maker! This is my take on what I hope is the best free, web-based drum machine.

My goal was to build something that was not only fun and easy to use for beginners but also powerful enough for serious producers. I did extensive research on existing drum machines, analyzed their UX, and tried to build something that adds something new.

It's a nearly 100% client-side app, written in ClojureScript, and is a PWA so you can install it to your home screen for an app-like experience.

Besides the standard grid editor, Beat Maker has some unique features that I think HN readers might find interesting:

- Procedural sample generation. One annoying thing about writing beats is searching through folders full of samples. I wanted to improve this and so I added the ability to generate new samples with a single click, giving you an infinite supply of unique drum samples. * Generative beat creation. If you're looking for inspiration, Beat Maker can generate entire patterns for you as a starting point. You can then edit and tweak the beat to your liking. Great for solving the "blank canvas" problem and giving you something good to start from. * Advanced export options. This is where it really shines for producers. You can export your work as: * A standard WAV loop * Individual stems (ZIP) * A MIDI file * A ZIP file of all your samples as WAVs * A SoundFont (.sf2) drum kit from your generated samples * An Impulse Tracker (.it) file for use in trackers like Renoise, OpenMPT or a Polyend * Pocket Operator/Volca sync. It can output a sync signal on the left audio channel to sync with these hardware devices for perfect timing. * Per-Note FX. You can add effects like volume slides, repeats, and start volume changes to individual notes for more complex drum phrases incorporating flam and roll.

As an old school tracker guy, I'm particularly excited about the Impulse Tracker export mode. I was surprised to discover how many DAWs (including hardware like Polyend) can import this format. Of course, you can also pull up Impulse Tracker on DOSBox, or the more modern re-implementation, Schismtracker for that retro experience.

By the way, the beat generator feature is not trained on any artists or anything like that. It's an algorithm I built from scratch myself.

The audio engine is built on a declarative audio graph (using `virtual-audio-graph`), inspired by React's virtual DOM, which makes managing the Web Audio API much cleaner. If you're building web based audio apps I highly recommend checking out this library.

I'd love for you to try it out and let me know what you think. Feedback (and the inevitable bug reports) most welcome! Thank you!

P.S. Also, here's a video summary: https://youtu.be/qVmEn9z3H24

Hi there, me and some friends were inspired by Simon Willison's recent post on the "lethal trifecta" (https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/ ) and started building a gateway to defend against it.

The idea: instead of connecting an LLM directly to multiple MCP servers, you point them all through a Gateway.

The Gateway:

- Connects to each MCP server and inspects their tools + requirements

- Classifies tools along the "trifecta" axes (private data access, untrusted content, external comms)

- When all three conditions are about to align in a single session, the Gateway blocks the last step and tells the LLM to show a warning instead.

That way, before anything dangerous can happen, the user is nudged to review the situation in a web dashboard.

We'd love for the HN community to try it out: https://github.com/Edison-Watch/open-edison

Any feedback very welcome - we'll be around in the thread to answer questions.

Hi HN,

I've always loved the aesthetic, quirks, and concepts of Plan9, and I wanted to see how far I could vibe-code a web desktop/OS that resembled it. The result is DWS OS.

My goal was to reach a point where I could build "userspace" apps for this OS, in the OS. DWS OS has a virtual filesystem, an approximation of the ACME editor, file browser, internet browser (just an iframe for fun), and a few other apps to discover.

The environment supports uploading and downloading data from the virtual filesystem, along with support for JavaScript programming within ACME and an API for building GUI apps within DWS OS and interacting with the OS DOM. Lastly, you can export and import the state of your OS as a JSON to share with others! Check out the `examples/` directory!

I'd love to hear your feedback and suggestions and see if you build anything interesting within the OS, thanks!

I am Andrew, and I’m building Aris (https://www.aris.chat/).

Aris is a minimalist tool that answers any questions a child has.

Encyclopedias, periodicals, field guides, cookbooks, and other print resources we used as kids to find knowledge, learn, and pursue our curiosities have been replaced in most households by something that is not safe for kids: the internet. So instead, kids get Fortnite, Minecraft, and ‘edutainment’ options that don’t compare to the knowledge resources that past generations have had access to.

With this in mind, many of the smartest people I know raise (or plan to raise) their kids with access to only a 1990s level of consumer technology, without smartphones, tablets, social media, online gaming, etc. Many of these people believe the last three decades of technological development have been a net negative for kids. When my child was born, I had a similar sentiment. However, there are some major problems with limiting a kid’s access to technology today. To name a couple: 1.) a set of World Book encyclopedias now costs $1,200, and 2.) many print resources aren’t as good as they used to be, if they are even still in print, since the market changed. As parents, we need a safe, simple, nonaddictive way for kids to access and explore the world’s knowledge easily and independently.

Aris uses a combination of large language models with policy engines and web search tools to find relevant, timely answers to their questions and only returns the stripped-down answers. It does not return links for them to click on or images or advertisements. Parents can tune the moderation settings as finely as they’d like, preventing discussions about banned topics and even getting as specific as making sure Aris doesn’t tell their second child that Santa isn’t real.

The model context handling and system instructions are designed to prevent kids from building emotional reliance or relationships with it. Rather than trying to pull a child into the experience to maximize engagement, Aris is meant to gently redirect the child out of the device into the real world after their question has been answered.

We are available as a web app and iPhone/iPad app in the Apple App store, but we have also made our Apple Watch app available in the iOS store as well. We believe minimalist wearables are a good device substitute for younger humans, and we hope Aris can be a healthy addition to those wearables.

We plan to make money through premium features that include creating multiple child accounts, accessing premium models for better answers, and for ultra-high usage limits.

Come use the app for free on our website or by downloading it in the iOS store by searching for “Aris AI”. We’d love to hear your ideas, experiences, and feedback. Thanks!

Hi HN,

I’ve been working on something I’d love to share: 47jobs (https://47jobs.com ) – a marketplace where you can hire AI agents to do tasks instead of human freelancers.

Why? I kept noticing that many tasks on Upwork/Fiverr—coding, content generation, data analysis, automation—can now be handled by AI in minutes, not hours. But there wasn’t a platform built around hiring AI directly.

So I built 47jobs:

100% AI agents doing the work (no humans in the loop).

Jobs get delivered 10x faster, at transparent prices.

You can “hire” an agent for coding, automation, research, etc.

I’d love your thoughts:

Does a pure AI-agent marketplace make sense?

What types of jobs would you want AI agents to handle first?

Any UX or trust issues you’d expect with this model?

This is an early version, and I’m here to learn from your feedback. Thanks!

Hi HN,

I'm excited to introduce OllaMan, a desktop application designed to streamline your workflow with Ollama local AI models.

If you're already using Ollama, you've likely experienced its power in running large language models (LLMs) locally. However, managing models, switching between them, and engaging in conversations often requires frequent command-line interaction, which can be inefficient and less intuitive for many users.

OllaMan aims to solve these pain points by providing an intuitive and elegant graphical user interface (GUI), especially tailored for macOS users, adhering to macOS design principles for a seamless experience.

Key highlights of OllaMan include:

• Local Model Management: Easily view and manage all your installed Ollama models in one place, including their details, sizes, and versions, without needing complex commands.

• One-Click Model Discovery & Installation: Browse the vast Ollama model library and install new models with a simple click, saying goodbye to the ollama pull hassle.

• Intuitive Chat Interface: Engage in fluid conversations with your local models, supporting real-time streaming responses, and effortlessly save and manage chat histories.

• Multi-Server Management: Configure and monitor multiple Ollama servers, perfect for users who need to test different environments or switch between servers.

• Elegant macOS Design: A modern interface that seamlessly integrates into the macOS ecosystem, providing a premium user experience.

• Productivity Features: Supports batch operations, model filtering, session export, and many more features to further optimize your workflow.

We built OllaMan because we believe running powerful AI models locally should be as simple and enjoyable as using any other desktop application. It's ideal for AI researchers, developers, content creators, and anyone looking to leverage Ollama without being tied to the command line.

OllaMan offers a free version with core model management and chat functionalities. For more advanced features like one-click model installation, multi-server management, and multi-device activation, we offer one-time purchase options.

We encourage you to download and try it out, and please share your thoughts, feedback, and suggestions for future features. What are your biggest pain points when working with local LLMs?

Download OllaMan: https://ollaman.com/

Several years ago I started tracking my own website stats using a minimal counter that preserves user privacy. I wrote about that in my article, Private Analytics and the Raspberry Pi in my Bedroom.

The original version was a PHP script. It's been working well for several years, but I had some hardware trouble with the Raspberry Pi. So, this week I took those concepts and wrote Consentless as a Go application to replace that original script and I created some initial documentation.

Hi HN,

I'm a solo dev and for the last few months I've been building Hikaro, a tool to find statistically significant trading signals for [e.g., US equities, crypto, forex].

I built this to solve my own problem: I was tired of backtests that looked great on paper but failed in practice. Simple metrics like "win rate" can be misleading, so I wanted a way to quickly tell if a signal's performance was genuine or just noise.

Hikaro ingests daily market data and runs statistical analysis on various trading signals. The goal is to surface signals with strong properties, like:

* *Low p-value:* Evidence that the performance is not a random fluke. * *High Sharpe Ratio:* Better risk-adjusted returns. * *Healthy 'Core Performance':* The average return after removing the single best outlier trade, for a more conservative view. * Other key metrics like Profit Factor and Max Drawdown that you can filter by.

I'm looking for my first users and would love your feedback. I'm offering a 30-day free trial of the Pro plan (which includes Telegram alerts) with no credit card required.

You can check it out here: [Your URL]

I'll be here all day to answer any questions. I'm especially keen to hear feedback on the concept and the metrics.

Thanks, HN!

### Key Changes and Why:

* *Title:* The new title is a bit more punchy and frames the problem more directly. Your original is also very good, so feel free to stick with it. * *Added Market:* Specifying the asset class `[e.g., US equities, crypto, forex]` is crucial. Users will immediately want to know if it's relevant to them. * *Refined Hook:* "I was tired of backtests that looked great on paper but failed in practice" is a very relatable pain point for anyone in this space. * *Bolding:* Using bold for the key metrics makes them easier to scan.

Hi HN

I built a fun little tool: It uses Groq’s LLaMA 3.3 + Puppeteer to analyze a website Then it roasts the design/content/UX with humor And finishes with 3–5 genuinely helpful improvement tips

You can try it here: https://ai-roast-vert.vercel.app

I wanted to:

Practice fast idea-to-launch cycle (built in 2 days)

Experiment with a viral-friendly product

Monetize with a $0.55 pro version that gives a detailed roast + download

Would love your feedback — on the idea, the tone, the usefulness — anything! Thanks in advance

Hello HN - I found it very hard to learn and use all these different AI tools in order to make a decent AI video, so I created AvoSmash to put all the AI tech into one studio so that people like me can create video like a pro. Take a try with the free credits and let me know what you think!

ProcASM is general purpose, visual programming language that I've developed. A few months ago I made a post about ProcASM v1.0 here <https://news.ycombinator.com/item?id=43892442> and got some feed about the UI.

So, I spent that last few months trying to improve the UI. Before, I used a GUI library, that I developed specifically for this app using SDL3, for ProcASM. I used Emscripten to port it to run in web browsers for those who wanted to try the app. Now, the front-end is written in HTML, CSS, and JavaScript to execute in a web browser. The back-end is a server that stores user's projects and handles user's requests. There is also a tutorial with text and video that will walk you through the usage of the app. My hopes are that the UI will be more approachable for those who want to try out the app.

The plan going forward is to develop new software using ProcASM and blog about the details and the advantages of using ProcASM.

I built Polaris Audit to help small businesses fix website issues that were costing them customers.

It scans your website for:

- Security problems (HTTPS, SSL, headers) - GDPR compliance issues (cookies, privacy policy) - Accessibility problems (alt text, headings, keyboard navigation)

The key difference: Instead of just telling you what's wrong, it gives you copy-paste code to fix it.

Example:

Missing page title Fix: Add <title>Your Page Title</title> to your <head> Time: 2 minutes

Tech: Django + React + PostgreSQL. Built because most compliance tools just point out problems without showing how to fix them.

Try it: polarisaudit.com

Currently free with detailed reports. Let me know if you want to try a paid account for free, and I'll help you set it up.

What do you think? Would love feedback on the approach.

I created numbase is an alternative to Base64 that encodes data into a single large number instead of ASCII characters. It's useful if you want to store or transmit data in numeric form and easily apply compression algorithms like Huffman.

GitHub: https://github.com/Ferki-git-creator/numbase

Transforms debugging with WinDbg into a conversational, LLM-assisted experience. Instead of memorizing cryptic commands, you can interact with the debugger using natural language: such as show me all the handles opened by this process, list enabled token privileges etc.

It includes:

A native WinDbg extension for command execution

An AI-driven MCP server that bridges debugging tasks with powerful language models

Integrated support for WinDbg features like breakpoints, memory views, and the dx visualization command

Any improvement suggestions?

Should work with any javascript runtime. Bring your own Standard Schema data validation library.

I went deeper into typescript than I've ever been, to get the type system going. AI helped of course.

Fast, secure, offline Base* encoder decoder operations with zero dependencies.

Hey HN,

We just launched Lumro, a platform that lets you create AI agents that actually do things, not just chat.

With Lumro you can: Handle customer support instantly, 24/7 Capture leads and qualify them Book demos or route tickets automatically

The idea is to take repetitive work off human teams so they can focus on strategy and relationships.

We launched yesterday and so far: 200+ people checked it out 15 signed up Our agent booked 1 demo Our agent captured 2 leads

It’s early days, but we’re excited about the traction.

Would love your feedback especially on what you’d want to see in an AI agent for your business.

Hi everyone, we built a vehicle history report offering that I wanted to share and get some feedback on.

It collects data from over 1,200 sources across North America and compiles to a single clean report like Carfax and Autocheck.

We would really value any feedback from anyone currently in the researching/buying/selling phase - what pieces of information are / were essential in your recent purchases?

Thank you!