Julian Collins

Matching obviously gets better with scale, but doesn't that create a really tough chicken-and-egg per campus? If Duke has 200 users and a smaller school has 12, those 12 students aren't getting useful matches. What's the minimum number per school before it starts working?

If you ever add mobile app support, that'd open up a huge use case. The gap between someone tapping 'Get' on the App Store and actually finishing onboarding is where most installs die, and there's no good way to simulate that flow right now. Desktop web is the right starting point but mobile onboarding is where the data gets really interesting.

I've been doing the exact loop you describe, screenshot on phone, export to Mac, edit in Figma, export again. Quick question: do you have plans to support App Store screenshot sizes? Going from a raw capture to something App Store ready without leaving the phone would save me a ton of time.

The MCP server integration is what makes this more than a nostalgia project. Everything in the terminal is about output and productivity, so building something in that space that's deliberately playful is a cool design decision. I'd actually want to see what happens when an agent tries to keep the pet alive during a long build.

How do you decide what goes into the AI reports vs what stays as raw data? I've been thinking about this with apps that deal with daily logging. There's a fine line between surfacing useful patterns and showing people so much analysis they stop trusting their own read on their numbers.

Did you consider building this as a share sheet extension or Shortcut instead of a full keyboard replacement? Custom keyboards on iOS have always had a rough time with the Full Access permission. Users get nervous granting it even when the app isn't collecting anything.

Since the task data is just a local JSON file on the Mac, have you thought about documenting the format so people can pipe tasks in from scripts? Something like `echo task | poptask add` from the terminal. Keeps the no-cloud thing going and opens it up to the developer crowd without touching the main UI.

The view tracking is the part that actually changes the dynamic. I've sent work over Google Drive and the worst part isn't bad feedback, it's wondering if they even opened it. Knowing what someone actually looked at before they respond gives you a completely different starting point for the conversation.

Three years on one product is serious commitment. I'm working on an iOS app and constantly have to resist the urge to add complexity just because I can. The fact that your filters work the same at every level, globally or inside a single list, is the kind of decision that only comes from living with something for that long.

The migration flow for existing subscribers is the part that caught my eye. Every other direct billing tool I've seen focuses on new signups, which means you're stuck running two systems indefinitely. Actually moving people off App Store billing is where the real margin recovery happens.

Smart call starting with slide videos and explainers. Those are the ones where the on-screen text basically IS the content. Quick question though, how does the editable text handle cases where the translated version is way longer than the original? Like English to German where labels can almost double in length. Does it auto-resize or does someone need to go in and adjust the layout?

The built-in cloud database is smart for getting prototypes out fast. But what happens if someone builds something real on this and then needs to migrate away? Can you export the data layer, or are you pretty much committed once you start building on Nativeline Cloud?

The interception approach is clever, way faster than spinning up a headless browser for every request. Have you thought about a batch endpoint where you can throw a list of URLs at it in one call? Anytime I've built a scraping pipeline for a project, the single-URL-at-a-time loop is where things get slow and annoying to manage.

Been working on an iOS app myself and getting people past that first-session hesitation is always the hardest part. The scenario structure probably helps a lot with that. Giving someone a specific thing to do ("order a coffee") is way less intimidating than just "go practice." Congrats on #1.