Hey everyone — I’m wavever, the developer behind Clipth.

I’ve been working on Clipth for quite a while because I wanted a clipboard manager that felt natural for keyboard-heavy work. Most of the time, I don’t want to stop and reach for the mouse just to find or paste something I copied earlier.

Clipth is a native macOS clipboard manager with:

• A keyboard-first Quick Paste panel
• Fast search, pinning, favorites, groups and tags
• Image, video and webpage previews
• List and grid layouts
• Sync support for iCloud, WebDAV, folders and S3
• QR code tools, OCR, Base64, image/video conversion and compression
• Custom Shell and JavaScript scripts
• A local MCP server for AI clients to work with clipboard data

The core version is open source: https://github.com/wavever/Clipth

The Pro version adds the advanced tools and automation features. It’s a one-time purchase, not a subscription: $10.99 for 2 Macs.

There’s also a launch price of $8.99 with code HELLOCLIPTH until August 26.

Website: https://clipth.app

I’d really appreciate feedback, especially from people who use keyboard-driven workflows on macOS. What would you want to see in a clipboard manager?

Everything on a Mac has quietly got heavier, including the apps whose whole job is to tell you how heavy everything is. A menu bar monitor runs every second the machine is on, and reading a few kernel counters twice a second is not hard work. It should not cost much.

AirStats is 10.6 MB.

It reads CPU, memory, GPU, network, disk, battery, temperature and top processes directly through Mach, IOKit, sysctl and CoreWLAN, and draws them in three places:

• the menu bar itself
• a panel that drops down when you click the status item
• an optional floating overlay you can leave on the desktop

Written in Native Swift, no Electron, no web view. One process, with no helper, no launch agent and no Dock icon.

Other menu bar monitors, like iStat Menus are more thorough than this and have years of polish behind them. Stats is the open-source one most people already know. However, AirStats is narrower on purpose. It covers fewer sensors and spends the effort on staying small and on the overlay, which is the part I use most.

There are no analytics and no telemetry, and readings never leave the machine. It makes two outbound requests, both visible in the source: an update check, and a public IP lookup that is off unless you turn it on. Policy at https://airstats.app/privacy

It is free, with no account and no subscription. macOS 14 or later, Apple silicon. Signed and notarized, so it opens with no warning.

• Site: https://airstats.app
• Source: https://github.com/byrencheema/airstats

If there is a feature you want or an issue you hit, feel free to reach out.

6 months ago I left a decade of fintech engineering in London to build Lyra solo ,a Mac writing assistant that fixes your writing as you type, in every app.

The problem I built it for: everyone I worked with had the same ritual. Type a message, re-read it four times, paste it into some tool, fix it, paste it back, tweak the tone, send. Multiply by 50 messages a day. And the tools that "help" all have the same flaws — they live in a separate window, they interrupt you with underlines while you're mid-thought, and they sand everyone down to the same polished-robot voice, whether you're messaging your boss or your best mate.

So Lyra works differently: you just type. Pause, and a small pill shows a better version — press Tab, done. No panel, no copy-paste, no clicks. It works in Gmail, Outlook, Slack, WhatsApp, your browser, wherever you write. And it learns how you write to each specific person, so your message to a client and your message to a friend don't come out sounding like the same press release.

Launched quietly in July. Show HN flagged my post. Reddit: silence. X: silence. Then one person paid ,a guy in France. I was thrilled for about an hour, which is how long it took him to uninstall: Lyra was freezing his Mac every minute.

Here's what that month taught me. I assumed the worst (memory leak), built instrumentation, ran controlled tests on identical hardware and my confident assumption was wrong. The real cause was a design decision I'd shipped without questioning. Fixed it, the freezing died, and he told me he'd reinstall ,he'd called Lyra "a really great product" while uninstalling it, which is the strangest compliment I've ever received.

Since then: fixed the biggest issues my ~30 early users hit, rebuilt the rough edges, made a proper demo. Relaunching today.

Mac only. 14-day full trial, free tier after, $9.99/mo for Pro.

Happy to answer anything , the debugging saga, why the first launch flopped, what one uninstall taught me that 30 signups didn't, or the unglamorous reality of marketing as a solo dev.

Try Lyra here Get Lyra

I'm an ex-Apple engineer who got sick of non-private, non-native meeting recording apps, so I built Biscotti: a private and free meeting recorder for macOS.

Free, no account, no subscription, no upsell. Source and download on GitHub: https://github.com/scosman/Biscotti

What it does

• 🔒 Private by design - recording, transcription, and AI all run on your Mac. Your data stays local.
• 🧠 Powerful AI summaries - automatic summaries, action items, meeting titles, and real speaker names.
• 🤖 No bots, any app - records Zoom, Teams, Meet, FaceTime, Slack huddles, or even an in-person conversation. No bot joins your call.
• 🗣️ Knows who said what - accurate transcripts, automatically split by speaker.
• 📅 Calendar-aware - sees your upcoming meetings, offers to start recording.
• ⏹️ Auto-stop - detects when your call ends and stops recording.
• 🎤 Voice isolation - captures your mic and everyone else as separate, clean channels. No echo.

Background

I spent years at Apple building the Photos app, Memories, the lock screen, and more. I wanted this to feel like it belongs on the Mac: Swift and SwiftUI, launches instantly, small on disk. Not a Electron app or a mediocre port of a Windows app.

Private AI == Local AI

With Biscotti, you still get state of the art AI for transcription and speaker identification, it just runs locally on your Mac. It works fully offline once models are downloaded. Your data lives on your Mac - you own and control it.

It uses Whisper for transcription, Gemma 4 for summaries, and Pyannote for Speaker ID. Each set up to run on Apple Silicon.

Requirements: Apple Silicon (M1+), macOS 15+, 16GB RAM recommended.

App details: Notarized, Free (no upsell), Private (zero data collection, all AI is local), Source available on Github (PolyForm Perimeter License), AI used in development (but not vibe/slop)

I'd love feedback! Happy to answer anything about the local model stack or the privacy design.

I built ImageBuddy, a macOS utility for cleaning up images: trim dead border space, remove the background, resize the image or canvas, add a watermark, vectorize to SVG. Drop an image, toggle what you need, save. You can edit a single image or set up a batch.

Notarized direct download or Mac App Store for $5: bendansby.com/apps/imagebuddy.html

Also available in a bundle of other apps: bendansby.com/apps/bundles.html

I built HeyZoku.
PH - https://www.producthunt.com/products/heyzoku/launches/heyzoku-hands-free-coding

As the models are getting much better at understanding us directly via voice, we should be working towards finding more ways to improve our productivity and do better work. And shouldn't be limited to just chat/keyboards to do our work. With how humans naturally communicate with each other, it should be the same way with the machines, a human shouldn't need to learn how a computer works - it should be natural.

So today I am launching HeyZoku - Hands Free Coding. So you can control all your coding agents just with your voice and complete your work without any need to continuously touch your keyboard.
I am using Nvidia Parakeet to make it work completely offline - so it can work completely offline. (It also has an optional bring your own openai-realtime key for more natural and multi language support.)

Open to any kind of questions or feedback and thank you for your support.
An upvote would mean a lot!

I’ve been building Tapquence, a visual automation app for macOS.

The idea is simple: repetitive work shouldn’t keep taking your time. Tapquence lets you build visual macros without code, combining actions such as opening apps and websites, typing text, clicking, scrolling, waiting and working with content on your screen.

It also includes ready-made Smart Automations. Presentation Mode can prepare your Mac for a meeting by checking your calendar, opening the relevant meeting app, launching your presentation and websites, silencing notifications and adjusting the volume. Routine Scout can observe a workflow you choose to perform and help identify repeated steps that could become an automation.

This short video shows the current version. It’s still evolving, so I’d genuinely appreciate feedback on what feels useful, what remains unclear and which automation you would want us to add next.

Tapquence is available directly from our website. The Mac download is signed and notarized by Apple.

https://tapquence.com

What repetitive task on your Mac would you most like to automate?

Most SSH clients on Mac are either Electron wrappers or generic terminal apps with SSH bolted on. I wanted something that felt like it belonged on macOS, so I spent 6 months developing my product solo.

"Native" gets thrown around a lot, so to be specific about what I mean: the whole app is Swift/SwiftUI, including the terminal emulator (not a libvterm fork, not xterm.js in a WebView) and the Mosh client (own implementation in Swift with a complete framebuffer mirror). No Electron, no web shell, no hidden Chromium.

A few things I'm proud of under the hood:

- SSH stack on top of Citadel + a fork of swift-nio-ssh with a window-adjust patch and a custom RSA path for legacy key formats some servers still hand out

- Own SOCKS5 server written in Swift (RFC 1928 + 1929 auth) for dynamic forwards, so `ssh -D` works without shelling out to /usr/bin/ssh

- Active tunnels (local / remote / dynamic) get snapshotted and restored on reconnect, so dropping Wi-Fi at a cafe doesn't kill your forwards

- Smart Reconnect that survives a cold launch, not just a network blip

- ProxyJump multihop with a runtime session graph

- Multi-host broadcast with a diff view across panes

- FIDO2 / sk-* hardware keys via libfido2

- SFTP File provider

- Built-in S3 client (AWS + R2 + B2 + Wasabi + MinIO + custom endpoints) with a real object browser, multipart uploads, versioning/lifecycle/CORS/encryption settings, presigned URLs and ACL/tag editors. Same window as your SSH sessions

- Keys live in the macOS Keychain (iCloud sync off by design), can be generated inside the Secure Enclave with Touch ID required on every signature, or imported from existing OpenSSH files. Passphrase-protected ones decrypted in-app

Solo project, so honest feedback on what feels off is genuinely useful. Happy to answer technical questions in the comments.

ProductHunt: https://www.producthunt.com/products/termy

Termy is now available in EAP: termyapp.com
Code MACOSAPPS50 is 50% off any plan for the first 25 purchases.

Hey everyone,

My name is Scott. I spent most of my career using Windows before switching to Mac so I could expand my development setup. I was impressed by the hardware, but years of Windows muscle memory made the transition more difficult than I expected.

I kept pressing the wrong shortcut keys and missed features such as taskbar window previews, clicking an active app to minimize it, and having a familiar launcher for finding applications.

I initially looked for apps that could solve each of those frustrations, but I would have needed several separate utilities. I eventually decided to build one unified and customizable solution called Winstrix.

What Winstrix does

Winstrix brings a collection of Windows-inspired workflow features to macOS while letting you choose exactly which ones you want to use, with the ability to disable the features you don't want.

Taskbar with live hover previews

Hover over an app icon to see live thumbnails of its open windows. You can select or close a specific window directly from the preview. Clicking an active app icon minimizes its window, and holding the icon minimizes all windows belonging to that app.

Start-style launcher

A dedicated launcher with unified search, pinned and recent applications, and an all-apps view for quickly finding anything installed on your Mac.

Windows shortcut mapping

Map familiar shortcuts such as Ctrl+C, Ctrl+V, Ctrl+Z, and Ctrl+Alt+Delete so you don't have to constantly fight years of Windows muscle memory.

Clipboard history and utilities

Searchable clipboard history with the ability to AirDrop copied items to your phone, along with a clock and calendar tray with customizable system shortcuts.

Desktop and menu bar controls

Quickly hide or reveal desktop icons and crowded menu bar icons without having to change several different macOS settings.

Screenshot and screen recording tools

Built-in screenshot capture, screen recording, and an integrated editor.

Linear mouse wheel scrolling

Removes the acceleration curve that can make a standard mouse wheel feel jumpy or inconsistent. Each wheel movement scrolls a predictable amount.

Fully modular

Every feature has its own independent switch. If you only want the taskbar and Windows shortcuts, you can turn everything else off. You can also disable Winstrix completely and return to the normal macOS desktop with one switch.

Winstrix is built entirely in native Swift without Electron. It remains lightweight, and all of its features run locally on the Mac. There is no account requirement, cloud connectivity, telemetry, or analytics tracking inside the app.

Privacy

Winstrix does not collect or transmit personal data or usage analytics. Your files and data stay on your Mac.

Privacy Policy: https://winstrix.app/privacy/

Apple notarization

Winstrix is signed and notarized by Apple and is distributed directly from the Winstrix website rather than through the Mac App Store.

There is a fully functional 14-day free trial for anyone who wants to try it out.

The price is $14.99 for a lifetime perpetual license with no subscription.

Website: https://winstrix.app/

I'd really appreciate honest feedback from anyone who gives it a try. I'm especially interested in hearing whether anything feels confusing at first, whether there are Windows behaviors you expected to see that are still missing, and which features you would realistically keep enabled every day.

When Apple first introduced the "liquid glass" effect, opinions were quite divided; some criticized its visibility, others admired its aesthetics, and some worried about battery life.

I remained fairly neutral, I found it interesting though.
To be honest, the Apple apps featuring this effect were quite buggy at launch.

Now that it has stabilized and the bugs have been ironed out, I’ve implemented it in my own app, and I think it looks great and modern. Here is the Repo
It’s just a custom theme (since I know many people don't like it), so it can easily be turned off.

Icons animate in response to certain changes, and the background, which looks as if viewed through a layer of glass and clear liquid, offers a pleasing, relaxing visual experience.

I haven't encountered significant issues with app performance or battery life, though I'm curious what others think. :)

PS: App is notarized

Leafy is a Mac app for looking up and learning words without leaving what you’re reading.

Press ⌥A, select a word anywhere on your screen, and Leafy reads it with OCR. It works with PDFs, ebooks, screenshots, video subtitles, and pretty much anything else on your screen.

Instead of giving you a long list of dictionary definitions, Leafy gives you the meaning of the word in the sentence you’re reading.

Every lookup is saved automatically, along with the sentence it came from. Over time, Leafy builds a personal vocabulary list from the things you actually read.

When you review a word, you see the original sentence again with the word hidden. No separate flashcard setup, no manually building word lists.

Currently:

• macOS 14+
• Menu bar app
• No account or ads
• 10 languages, including Japanese, Chinese and Arabic
• Export to JSON, CSV, Excel, PDF or Anki
• AI-powered lookups
• Your saved words and data stay on your Mac
• Usage statistics can be turned off
• Free during beta
• leafyapp.uk

It’s still in beta, and I’m mainly looking for feedback on how people actually use it.

Hi there,
I'm Maciej, a DevOps engineer building this as my own daily driver. I develop it with AI assistance, TDD driven and with mutation testing.

I posted here a few months back around v0.7.2. It's now at v0.8.39 (signed and notarized) and I've been shipping almost daily, mostly from user reports. The short version of what's new:

• Saved workspaces with different panel layouts.
• Multiple windows.
• Search was rebuilt with its own index and is faster than light :)
• Preview now handles 235 file types.
• File panel columns are configurable per pane (show/hide, reorder, resize).
• Sidebar is fully editable.
• The newest two (Go to folder on CMD+SHIFT+G and settings search) just shipped.

Pricing after the feedback last time: free tier for daily use, Pro is 15USD one time. No subscription, no telemetry, no data processing.

Download: https://salamander.codesurfer.ch/
Feedback: https://feedback.codesurfer.ch/

For privacy policy see https://salamander.codesurfer.ch/privacy.html.

Hey everyone, I’ve been working on ConsoleWise, a native macOS app for PS5 Remote Play and PlayStation cloud gaming.

The goal is simple: make it easier to play your PlayStation games from your Mac, whether your PS5 is in another room or you’re away from home.

A few things I focused on:

• Play your PS5 when you’re not home Connect over the internet and stream from your own console while traveling, at work, or away from your TV.
• Low-latency Remote Play ConsoleWise is built around responsive streaming, with adjustable resolution, FPS, bitrate, and video settings so you can tune the experience for your network.
• Cloud gaming without owning a PS5 If you have PlayStation Plus Premium and cloud streaming is available in your region, you can browse and stream supported cloud games directly from your Mac.
• Use your keyboard or DualSense Play with a DualSense controller, or map keyboard keys to controller actions when that’s more convenient.

It supports local and internet PS5 Remote Play, PlayStation cloud game streaming, 1080p streaming, 2K/4K local upscaling for Remote Play, custom bitrate controls, 30/60 FPS options, keyboard mappings, and DualSense haptics/adaptive trigger settings.

It’s available here:
https://consolewise.aowlly.com/

Clicking "View Pro Details" will take you to Gumroad.

Use code J3DU8RK for 50% OFF.

I’d love feedback from Mac + PlayStation users, especially around latency, setup, and what would make this more useful than the official client.

Enjoy!

I stepped away from programming about 6 years ago to take care of my family and never really went back to it.

Recently I came across Claude Code and just wanted to try it out.

At the same time, my husband works 8–10 hours a day in front of a screen. So I thought I’d try building something for him.

I ended up making a small Mac app that gives simple reminders during long work sessions:

• blink / eye breaks

• posture checks

• short movement breaks

• optional focus timers

Nothing fancy, just something that runs quietly and nudges you once in a while without being annoying.

Honestly, I was surprised I could even build this after being away from coding for so long. Claude helped a lot, especially with figuring things out as I went.

The app is a paid app right now. If you do try it and run into issues, feel free to comment here. I’d really appreciate the feedback.

Here the app

BlinkAway: Micro breaks

Looking forward to building something more complex next 🙂

With my PhD research experience, I built Atelier: Scholarly Workspace around a problem I kept encountering in research:
The difference is not whether research is linked or cited. It is whether scholarly judgment itself is preserved.

Obsidian connects notes. But a link alone does not make the relationship, rationale, or supporting passage explicit.

NotebookLM answers from sources and shows citations. But a cited answer is not, by itself, a durable researcher-authored judgment.

Atelier: Scholarly Workspace is built to preserve all three: the judgment, why you made it, and the exact place in the source that supports it.

Atelier: Scholarly Workspace for Research & Deep Reading

Hi! My Downloads folder had been left unkept for a very long time. I started cleaning it up with Finder but it was quite cumbersome so I started creating a simple app to help me filter out and delete or move the files in the folder.

It genuinely helped me clean out my Downloads folder and as I thought of more features I wanted in a file manager I started to add them.

Features such as: Fuzzy go to folder/file where you only need to type a few letters of the path to get suggestions and jump to the correct path instantly. Hover over an image or PDF file to preview it instantly. Command palette to find actions fast. Dual-pane view and tabs that remember their own filters.

As I used the app I added even more features that I wanted: Batch rename with regex support and presets for common operations. Converting HEIC/WebP/AVIF image files to JPG/PNG. Sidebar, single-click open, keyboard controls and many other small quality of life features I discovered while using it. Since it's a file manager I put real work into safe file operations, and the only network request it makes is an anonymous version check.

WhimFiles has a launch price of $19.99 and a fully-featured 30-day trial.

The trial can be downloaded here: https://whimfiles.com

I am Ighor, the developer of App Archiver.

I kept running out of space on my Mac and became frustrated with the usual solutions.

Most cleaner apps create space in one of two ways.

• They delete apps, downloads, or other data
• They remove caches that applications create again

Deleting data provides lasting savings because the data is gone. Clearing caches can show an impressive number immediately, but those caches often return as soon as you continue using your apps.

That is not the kind of storage management I wanted.

I like downloading Mac apps and keeping them for later. If I delete an app, I may forget it existed. I wanted to reclaim the space without losing the app, its familiar location, or the ability to find it again.

So I built App Archiver.

Real and lasting space savings

App Archiver reduces the space used by the apps themselves.

It does not pretend that repeatedly deleting temporary files is a permanent solution. It does not delete your documents, settings, or app data.

You can choose what happens to each app.

• Archive it in place
• Replace a rarely used app with a lightweight, restorable version.
• Optimize it
• Reduce its size while keeping it installed and ready to use.
• Back it up
• Create a standalone ZIP or DMG copy in any folder or external drive.

An archived app remains visible in Applications, Spotlight, Raycast, Alfred, and other launchers. When you need it again, double-click the same icon and App Archiver restores it.

The reclaimed space does not gradually disappear because a cache grew back. The full app remains archived until you decide to restore it.

No cleanup theater

I am not claiming that caches should never be cleared. Cache cleaning can help with troubleshooting and sometimes recover temporary space.

But repeatedly deleting files that return is not a lasting storage strategy.

App Archiver solves a different problem. It gives large and rarely used apps a smaller form while keeping them accessible. You get meaningful space back without throwing away software you may want later.

Real-world test

I tested Android Studio on an M1 Max using a single compression thread.

Result	Final size	Space saved	Reduction	Time
Original	3.48 GB	-	-	-
Optimized and runnable	1.76 GB	1.72 GB	49.4%	1m 50s
Archived	1.53 GB	1.95 GB	56.0%	3m 40s
Backup DMG	1.31 GB	2.17 GB	62.4%	4m 17s

The optimized version remains installed and runnable. The archived version stays in place and can be restored when needed. The DMG provides the greatest saving as a standalone backup.

These times include both compression and verification of every compressed file, not just the compression step.

This test intentionally used one CPU thread. I may add optional multithreaded compression in a future update to make processing considerably faster on multicore Macs.

Privacy

App Archiver works entirely offline. It has no account, sends no analytics, and makes no network requests.

One current limitation is that some apps are labeled Protected. Optimizing them requires additional system privileges, and I am waiting for Apple's permission before enabling that capability.

App Archiver requires macOS 12 or later and is available on the Mac App Store.

Download App Archiver on the Mac App Store

Visit the App Archiver website

Read the privacy policy

I would genuinely like to hear what you think about this approach to Mac storage management.

Also check out my other apps https://reverseeverything.com/apps

For the last year I've been building a writing assistant for macOS Lyra

At first, I did what everyone else does.

"Rewrite this."
"Make it friendlier."
"Make it more professional."

It worked.

But something felt off.

ChatGPT assumes you have one writing style. You don't.

I message my co-founder differently from recruiters.
I email my accountant differently from my wife.

The relationship changes the writing more than the tone does.

So I scrapped a big part of what I'd built.

Instead of learning my writing style, the app learns how I write to each individual person. Every accepted rewrite becomes another example for that relationship. After a while, messages to different people genuinely start sounding different.

The entire thing runs locally on Apple Silicon because I wasn't comfortable uploading years of private conversations just to improve my emails.

It's still far from perfect.

The local models occasionally miss difficult typos.
Learning a person's style takes a couple dozen accepted rewrites before it becomes noticeable.
And I'm a solo developer, so every support email comes to me.

But it finally feels like I'm building the product I originally imagined, instead of another AI rewriter.

What do you guys think ? Should AI learn you, or the way you're different with different people?

Lyra