-
-
See it, share it, it's pinned. Send any game clip straight from your feed and it lands right on your shelf.
-
Homepage works as a central hub where you see your own game acitivies as well as your friends, with a quick overview of your main path.
-
Discover page works as a way to discover new games and gamers, as well as lists you can find to see what other people are playing.
-
Search page where you can find games, players, and lists.
-
Detailed game information page that contains detailed descriptions, visuals, reviews, and the ability to manage that game in your lists.
-
Inventory page, one page solution to all your lists and paths.
-
Profile page; see your friend's profile, taste match, socials, and their lists. Match uses top 4 top games of that player to match with you.
-
Pinboxd app icon
Inspiration
We kept forgetting the games we actually wanted to play.
A clip shows up on TikTok at midnight, the gameplay looks incredible, and for a few seconds we're genuinely hooked. Then the feed moves on and the name is gone, buried under fifty other clips we'll never open again, and by morning we can't even remember what it was called to look it up.
It happens constantly, and nothing out there really solves it. Backlog apps assume you already know the name. They start after the exciting part is already over.
Pinboxd starts right at the clip. Share it, we catch the game, it's on your shelf before you've even scrolled past. We built this because we wanted it for ourselves first, and because a bucket list for games should feel like something you're proud of, not another list you're avoiding.
What it does
Share a TikTok, Reel, or Short into Pinboxd and we match it to the real game before the clip disappears. The game info behind every pin, covers, details, stays current too, so nothing on your shelf feels out of date.
From there you save it, play it, rate it, and share it. Path is your current, ordered list, it's meant to answer "what am I actually playing right now," not turn into another pile you never open again. Lists are unlimited and unordered. You get four by default, Dropped, Wishlist, Favourite, and Playing, and you can build your own on top, like a list just for games you're playing with friends.
Rate what you finish, write a quick review, export a share card when you want to show it off. Then there's the part people keep coming back for: taste. The moment you land on someone else's profile, you're checking for overlap without even thinking about it, it's the same pull that makes people scroll a stranger's Letterboxd looking for shared favourites. Your own top 4 sit right there for others to check too, and open one of their lists and we go further, showing you the percentage you've already played, so a friend's Wishlist turns into its own way to find what to play next, or a total stranger's turns into "wait, we like all the same things." If someone's added their Discord or other socials to their profile, that's right there as well, so finding someone whose taste matches yours can actually turn into talking to them.
How we built it
Flutter, iOS and Android, one codebase, clean architecture underneath.
Share-to-pin is fully native on both platforms so it works right from the share sheet, not a slow reopen of the app. Matching a messy caption to the right game runs on a lightweight language model tuned for exactly this one job, so it stays fast and cheap enough to run on every single pin. RevenueCat runs Pro end to end. We built this in public the whole way through Shipaton, showing people the ugly parts too, not just the finished screens.
Challenges we ran into
Getting share-to-pin to feel native took longer than we expected. TikTok, YouTube, and Instagram each hand you something different, and getting one dialog to handle all of it, fully in Pinboxd's theme, took real trial and error.
Matching was its own problem. A caption is never a clean game title, and running a full-sized model on every share would have been slow and expensive at any real scale. Language is exactly what these models are good at, so we used a lightweight one built specifically for this job, fast enough to feel instant and cheap enough to run on every pin without a second thought.
The icons mattered more than we expected too. A lot of us grew up on games with that exact chunky, pixel look, and we wanted Pinboxd to carry some of that nostalgia while still feeling like a modern app. None of us actually knew how to draw pixel art when we started, so we had to learn it first, then spent two weeks hand-drawing every icon in Figma, paired with a font called Monocraft that has the same retro feel. It reshaped the whole visual language of the app, but it made Pinboxd feel like ours.
The real risk the whole time was building another gray backlog list with nicer colors. Path, the Favourite match, share cards, all of it exists to keep that from happening.
Accomplishments that we're proud of
Share-to-pin actually working the way we pictured it before we knew if we could build it.
A lightweight AI model that matches a messy clip caption to the right game in under a second, without the cost of running something heavier on every single share.
A pixel art icon set we drew ourselves, not generated, not bought, carrying real nostalgia for the games a lot of us grew up on while still feeling like a modern app.
The Favourite match turning into something people genuinely go check other profiles for.
Shipping a real, working product inside the Shipaton window, not a demo.
What we learned
The first second matters more than anything else in the app. If share-to-pin doesn't work, nothing downstream matters.
People care more about comparing lists than we expected. We almost cut the Favourite match early on.
Building in public kept us honest. It's harder to quietly cut the hard part when people are actually watching you build it.
We spent real time studying apps like Letterboxd, Fable, Shelf, Backlogged, Goodreads, StoryGraph, and Serializd before writing a line of code, and it changed how we thought about what a "list" app should actually feel like.
Picking the right AI model mattered more than picking the biggest one. We had to learn what kind of model actually fit matching a messy caption to a game, not just grab whatever was newest.
None of us could draw pixel art going in. We had to actually learn it, and that turned out to be as real a skill investment as anything we wrote in code.
What's next for Pinboxd
v1 is the pin. Next is making the crew feel like a real crew, deeper library connections, and carrying the pixel art further into the app.
We're also looking at bringing films and shows into Pinboxd, the same instinct that makes you want to save a game from a clip applies just as much to a movie or a series you stumble across.
And we want profiles to feel like someone's, not just a template. Profile themes that carry through into your lists and reviews, so reading someone's review or scrolling their shelf feels like it's actually theirs. People should be able to express themselves here, and that's exactly what helps the right people find each other in the first place.
The pin from a clip was just the start. What we actually want is for Pinboxd to be the place gamers reach for when they want to show off their catalogue, the same way people already reach for Letterboxd or Goodreads. We're proud of what we've built so far, and we're not slowing down after October 1st, this is the beginning of something a lot bigger than a hackathon entry.
Built With
- admob
- dart
- digitalocean
- express.js
- figma
- flutter
- github
- next.js
- onesignal
- postgresql
- revenuecat
- typescript


Log in or sign up for Devpost to join the conversation.