Plánované datum vydání: 3. kvartál 2026

Informace o hře

Brace for Dolphins! is a fast-paced 2D, top-down roguelite strategy game where wars are won through speed, pressure, and hard choices—not perfect planning.

You play as an ancient AI, reactivated for one purpose: stop the expansion of an evolved dolphin empire that is rapidly conquering the galaxy. The enemy is relentless, adaptive, and always advancing. You don’t have the luxury of time.

Each run begins on a procedurally generated strategic map made up of contested nodes divided by a living frontline. Both sides deploy generals, fortify positions, and decide where to push next. However, each turn you can personally command only one battle. Every other conflict is simulated, with victories, defeats, and retreats dynamically reshaping the frontline and the future of your campaign.

When you choose a battle to fight, you crash-land on a contested planet and enter a short, intense real-time encounter. You must mine resources, build automation chains, and feed your weapons while under constant attack. There is no safe optimization phase—every second spent improving your base is a second the enemy is firing at you. Battles are designed to last 5–10 minutes, emphasizing rapid adaptation and decisive action.

Winning battles generates Momentum, which can be spent to upgrade individual generals with powerful traits that influence future simulations and fights. Retreating costs Momentum, and if a general dies, they are lost forever.

Defeat, however, is not the end. Every loss or retreat grants Intelligence Points that persist between runs. These points can be used to upgrade your entire empire, unlock new capabilities, and create new generals to help stabilize the frontline in future attempts.

Inspired by Factorio, FTL, and Dome Keeper, Brace for Dolphins! is about making the right decision under pressure. There is no perfect build—only the next move.

Informace o obsahu vytvářeném AI

Jak využití obsahu vytvářeného AI v tomto produktu popisují jeho vývojáři:

Generative AI was used in this project.