Slime Sanctum: Last Stand

Title
Lobby
3D Art
3D Dungeon Design
NoesisUI
AI NPC
Navigation
World Streamming
3D Modeling

Inspiration

WorldID: 888108847717337

Slime Sanctum: Last Stand was inspired by the fantasy of a desperate final stand in a single, iconic arena. We wanted the intensity of co-op horde defense, the build variety of roguelikes, and the immediate readability of mobile/web action games — all focused into one dark, dungeon-like sanctum overrun by cute but terrifying slimes.

What it does

Slime Sanctum: Last Stand is a co-op roguelike defense game where up to four Wardens protect a central Sanctum Core from endless waves of corrupted slimes. Players earn XP, level up, and draft randomized skills to create unique builds on the fly each run. The challenge is simple but deep: hold multiple lanes, react to new slime types, gamble on risky event rooms, and see how far your squad can push the wave counter before the Core finally falls.

How we built it

We started from a single constraint: one map, one core, many runs. From there, we built a modular sanctum layout with a bright central Core room and darker outer rooms that double as spawn points and event zones. On top of that, we implemented:

A wave manager that spawns different slime types and scales difficulty.

A light-weight XP and level-up system with a 3-card skill draft UI.

Simple but expressive combat (melee, ranged, magic) tuned for mobile/web inputs.

Clear VFX/SFX for hits, Core damage, and wave transitions to keep the battlefield readable.

We iterated quickly on pacing and layout, continuously testing how short sessions (about 8–12 minutes) feel when played in a loop.

Challenges we ran into

Balancing clarity and chaos was our biggest challenge. In a horde defense game, the screen can easily become unreadable: too many slimes, too many effects, and players lose track of the Core. We had to constantly adjust enemy counts, movement speeds, and VFX intensity to keep the action satisfying but legible.

We also wrestled with roguelike randomness vs. player agency. If skill drafts are too random, runs feel unfair; if they’re too predictable, they feel stale. Finding the right mix of class-leaning skills and common passives took multiple rounds of iteration.

Accomplishments that we're proud of

We’re proud that we managed to squeeze a lot of depth out of a single, iconic map instead of relying on multiple levels. The sanctum layout, with its central Core and outer threat rooms, naturally creates co-op roles and rotations.

We’re also happy with how the on-the-fly build system turned out. Even with a relatively small initial skill pool, runs already feel distinct: one session you’re a frontline cleave monster, another you’re a support mage shielding the Core and slowing everything down.

Finally, we’re proud of the overall game feel: the squishy hits on slimes, the escalating slime corruption visuals, and the “oh no, hold the line!” moments when the Core HP is hanging by a thread.

What we learned

We learned that tight constraints — one map, short sessions, small skill pool — can actually make design decisions clearer and faster. Focusing on a single space forced us to think deeply about sightlines, lanes, and co-op positioning instead of spreading effort across multiple levels.

We also reinforced how important readability and feedback are for action games, especially on mobile/web: clear telegraphs, distinct enemy silhouettes, and strong audio/visual cues made more difference than any single “fancy” feature. And on the systems side, we learned how to structure roguelike progression so players feel both surprised and in control.