GuildIt - A Little Fantasy, A Lot of Community

Inspiration

We wanted to build something fun; a game with a little bit of fantasy in it, classes, points, a reason to log back in tomorrow. But we also wanted it to mean something. Reddit is full of small, everyday kindness: someone welcoming a new user, someone stepping in to defend a person getting piled on, someone digging up a hidden gem of a post nobody noticed yet.

So we asked: what if a game could turn that everyday kindness into the actual gameplay? Not karma. Not a leaderboard for its own sake. A real community game, where being helpful, protective, curious, or creative is the whole point; wrapped in just enough fantasy (classes, guilds, duels) to make it feel like a game and not a chore.

That's GuildIt.

What it does

Pick a class. Each one is built around a way of showing up for people:

  • Ranger: finds the good stuff nobody's noticed yet
  • Mender: shows up for people having a hard time
  • Warder: steps in when someone's being treated unfairly
  • Weaver: starts the conversation everyone ends up joining

You earn points just by being a good part of the community. Post, comment, be kind, be helpful; the game notices and rewards it. No waiting.

Once a day, you can take on a Trial: a real question tied to your class. A Ranger has to spot a pattern. A Mender has to pick the kindest reply. A Warder has to make a real judgment call. Get it right, get full points. The answer is never something you can guess or cheat; it only lives on the server.

Some achievements have to be earned for real. A few of the biggest rewards only unlock once Reddit itself confirms your post or comment actually did something; really got upvoted, really sparked a real conversation. No shortcuts on those.

Guilds only work if they're balanced. A guild with all four classes scores at full strength. Missing a class costs you. That one rule does something nice on its own; it turns "we need one more player" into "we need a Mender," and suddenly people are actually recruiting each other, for real reasons.

Duels, without anyone needing to be online at the same time. Challenge someone. Whenever they accept- now, tonight, tomorrow- the whole battle plays out instantly behind the scenes. Then you both get to watch it happen, animated, like a real card battle, whenever you open the app. Nobody waits around.

How we built it

GuildIt is built on Reddit's Devvit platform, using Devvit Web; a real backend (Hono) talking to a React front end, with Phaser powering the duel animations. Points and guild data live in Redis. Real Reddit events (posts, comments) feed points instantly. A few actions get double-checked against real Reddit data before they pay out, so nothing can be faked.

Challenges we ran into

Time. We started later than we wanted to, so a lot of this got built fast and under real pressure, which meant making sharp calls about what mattered most and cutting anything that wasn't essential to the core loop.

Reddit doesn't tell you everything. We wanted points for things like upvotes received, awards given, and saved posts, but Reddit doesn't actually give developers that information. So a handful of ideas we really liked (UPVOTE_RECEIVED, DOWNVOTE_RECEIVED, AWARD_RECEIVED, CONTENT_SAVED, QUALITY_COMMENT, a Ranger "discover" event) simply aren't possible right now — not because we didn't build them, but because the platform has no way to tell an app that any of that happened. We designed around the honest limit instead of faking it.

Making the "trial" actually mean something. Early on, tapping a card just gave you points; no real decision involved. We caught that it felt hollow and rebuilt it as a real question with a real, server-checked answer. Small fix, but it's the difference between a game and a button.

Keeping two players in sync without either of them waiting. Duels needed to feel alive even though nobody's ever really "online together" on Reddit. Getting that to feel instant, not laggy, took real trial and error.

Accomplishments we're proud of

  • Built a progression system where every point is designed to trace back to meaningful community participation.
  • Created four unique classes with distinct identities that encourage different ways of contributing to Reddit communities.
  • Designed a guild system where cooperation isn't a suggestion—the scoring rewards balanced guilds with all four classes.
  • Built a secure, server-validated Trial system that rewards thoughtful decisions instead of repetitive actions.
  • Established the technical foundation for asynchronous duels and guild raids, allowing the game to grow beyond the hackathon.
  • Kept the fantasy elements fun while grounding the gameplay in real community values.

What we learned

Building GuildIt taught us a lot about designing games for Reddit and working within the platform's constraints.

  • Design around the platform, not against it. Reddit doesn't expose every event we'd like (such as vote events, awards, or saved posts), so we learned to build mechanics around the APIs that actually exist instead of relying on assumptions.
  • Meaningful progression beats meaningless rewards. Players enjoy earning points more when every reward is tied to a real action instead of simply pressing a button.
  • Simple mechanics are often stronger than complex ones. Reducing the number of classes to four gave each class a clear identity and made guild composition easier to understand.
  • Server-side validation matters. Sensitive actions, like daily Trials, should always be verified on the backend to prevent cheating and keep competition fair.
  • Asynchronous gameplay fits Reddit naturally. Not every multiplayer experience requires players to be online together. Designing around asynchronous interactions opens up interesting possibilities for Reddit communities.
  • Community-driven games require careful balancing. Rewarding helpful behavior without encouraging spam or abuse requires thoughtful scoring and validation.
  • Shipping an MVP means making difficult decisions. We learned to prioritize the core gameplay loop and postpone ambitious features that couldn't be polished in time for the hackathon.

What's next for GuildIt

GuildIt was intentionally built as a solid foundation first. Because we started the hackathon later than planned, we focused on delivering the strongest possible core experience rather than rushing unfinished features. While the core progression, classes, guilds, and Trials are functional, several larger features were only partially implemented due to the limited time available.

Our roadmap includes:

  • Complete the Duel system as a fast-paced, asynchronous social card game inspired by classic trading card games like Yu-Gi-Oh!. Instead of battling with monsters, players will play social action cards representing discovery, creativity, support, and protection. Each class will have its own unique cards, abilities, and strategies, creating deep tactical gameplay while staying true to GuildIt's community-focused philosophy.
  • Finish Guild Raids, where balanced guilds combine the unique strengths of all four classes to overcome large-scale community challenges. Success will depend on cooperation, card synergy, and having a complete guild composition rather than individual power.
  • Expand the collectible card system with new cards, class-specific abilities, combos, and seasonal unlocks.
  • Add weekly world events that encourage entire communities to work toward shared objectives.
  • Introduce seasonal rankings, guild achievements, and cosmetic rewards for long-term progression.
  • Create cooperative guild objectives that reward teamwork, coordination, and positive community participation.
  • Expand class progression with additional lore, unlockable abilities, and new strategic playstyles.

GuildIt today is the beginning of a much bigger social experience. We focused on building a strong, honest core that can continue growing well beyond the hackathon.

Built With

Share this project:

Updates

Submission history