Paul Renard 🦊

Hey Product Hunt! 👋 I built Squish because I was tired of emailing huge photos and watching my iCloud fill up with 5MB screenshots. The idea is dead simple: pick photos from your camera roll or any app via the share sheet, choose a quality level, and get compressed copies instantly. Everything runs 100% on-device — no uploads, no cloud, no waiting. What makes it different: • Works right from...

Love the simplicity here. One hotkey, no config, just works. That's the kind of Mac utility I respect. Does it play nicely with Stage Manager, or is it purely its own overlay approach?

Embedding audio directly in the image file is clever. So many travel memories lose context over time, and a voice note captures way more emotion than a caption. How does the audio embedding work technically? Is it stored in EXIF metadata, or a custom container?

Inline AI without context switching is a huge productivity win. The voice dictation angle is smart too. Is this using the system accessibility layer to inject text, or does it work through a custom keyboard/input method?

This is solving a real problem. I've been wary of handing credentials to AI tools, and the encrypted vault approach feels like the right architecture. Curious: does the vault support hardware key unlock (like YubiKey or Touch ID), or is it passphrase-only for now?

"Describe a mod in plain English → generate → reload → play" is a killer loop. You've essentially turned modding from a developer activity into a creative one, which massively expands who can participate. How do you handle mod conflicts when someone layers multiple AI-generated mods on the same server?

The share sheet → auto-organize flow is exactly right. Most "save it later" apps fail because they create a second inbox you never revisit. Does the AI categorization run on-device or does content get sent to a server? For a "save everything" app, the privacy model matters a lot.

This fills a real gap ! I've spent embarrassing amounts of time in generic video editors just to make a 30-second App Store preview. The "built specifically for product demos" framing is smart because the workflow really is different from general video editing. Do you support direct export to App Store screenshot/preview dimensions?

This is hilariously genius. The accept/reject/scroll loop with AI coding tools really is mostly binary decisions, mapping that to a gamepad is a surprisingly natural fit. Have you considered adding haptic feedback patterns for different event types (build succeeded vs. failed, new suggestion ready)?

Love the "drop a pin" UX for location reminders, it's one of those things that should've existed natively years ago. The map-first approach feels way more intuitive than typing an address into a text field. Curious how you handle geofence accuracy in dense urban areas where GPS can bounce around a lot?

Love the idea of metrics living on widgets instead of yet another dashboard tab. As an iOS dev, I appreciate when people build for the widget surface properly — most just shrink their main UI. What's been the most requested integration so far?

The "weekly relaunch" feature is a smart mechanic — one day of visibility is often not enough, especially for indie products that don't have a big audience on day one. How do you prevent it from becoming noisy though? Curation quality is what makes or breaks discovery platforms.

The concept of voice → polished text (not just transcription) feels like the right abstraction. Most dictation tools stop at raw transcription and leave you with a mess to edit. How much does it preserve your original voice vs. restructuring into "proper" writing?

This is the kind of native-first approach I wish more developers took — building on top of Ghostty instead of wrapping a web view. iCloud sync for SSH configs is a nice touch too. How's the performance on iPad compared to Mac?

Love the on-device approach — privacy-first is the right call for something as personal as a journal. Curious: how did you handle the balance between keeping everything local and still offering useful AI features? That's a trade-off I've been thinking about for my own iOS apps.