Tetra Dungeon

Deckbuilding roguelikes and Tetris have both been done to death individually, which makes Tetra Dungeon's central idea land as genuinely fresh: cards aren't just abstract effects, they're Tetris-shaped blocks that occupy space on a 5x5 battlefield grid, and how you rotate and place them determines whether an attack sweeps a full row, a column, or lands on a single target. Reviewers writing in Chinese, English, Japanese, Turkish and Russian all independently reach for the same word — novel — to describe this combination, and at a price under ten dollars, that novelty alone has carried the game to a Positive rating. What's harder to ignore is a specific, detailed criticism from an experienced genre fan arguing the roguelike structure wrapped around that novel idea doesn't hold up under real scrutiny.

Cards that are also puzzle pieces

The three-class structure gives the core hook real range. A swordsman uses straight-line blade cards for row-based attacks; a mage combines elemental spell shapes for flexible row, column, or single-target damage; a puppet master plays closer to tower defense, continuously repositioning summoned units across the grid rather than attacking directly. One detailed Chinese-language review lays out the appeal precisely: weapons deal immediate damage based on shape and placement, spells trigger row or column effects instantly, and summoned minions act every turn from wherever they're placed — meaning victory comes from maximising how these three card types interact spatially rather than simply accumulating the biggest numbers. Reviewers who enjoy this loop describe genuinely satisfying moments discovering a combo that "one-shots" a boss once fully assembled, and multiple reviewers across different languages compare the underlying rush directly to Slay the Spire.

A serious, specific challenge to the "roguelike" label

The most substantive criticism in the sample comes from a reviewer who identifies as an experienced card-game player, having spent time with Slay the Spire, Inscryption, Dicey Dungeons and Balatro. Their core argument, translated, is structural rather than emotional: the game's business and event nodes, and the sequence of combat encounters within a run, are largely fixed rather than genuinely randomised, and of the three bosses, only the first has any real variation — the other two are always the same encounter in the same order. They estimate that a player already fluent in the genre's conventions could clear all three classes and understand essentially the entire strategic depth of the game within about two hours, at which point, in their telling, refunding becomes a live option rather than continuing to a natural stopping point. That's a pointed claim, and it's worth taking seriously specifically because it comes from someone evaluating the game against real genre benchmarks rather than judging it in isolation.

Short runs, by design and by necessity

A second, related reviewer — this one with over seventeen hours logged, clearly someone who stuck around — makes a more measured version of the same point. They describe the game as, in their words, closer to a "小品游戏" — a light, vignette-scale release — rather than a fully realised roguelike, noting each run contains only nine combat encounters total (two regular fights and one boss per zone, three zones) and typically runs about twenty minutes. They're careful to frame this as an honest assessment of scope rather than outright condemnation, but the throughline connects directly to the more pointed criticism above: this is a small, tightly-scoped game, and reviewers disagree mainly on whether that scope was communicated honestly by the "roguelike" tag rather than on whether the core combat itself is any good.

A specific bug, and uneven balance

Beyond structural concerns, one reviewer reports a reproducible technical problem: killing a specific boss — described as a "Male Fire Dragon" — before it gets to attack can freeze the game entirely, requiring a return to the main menu, with the same softlock recurring on retry unless the player deliberately takes damage from that boss first before finishing it off. A second, unrelated performance issue is reported with a card called "Ghost Dagger." Separately, several reviewers describe class balance as rough: a particular sword-and-assistant combination can trivialise encounters, the mage is described by more than one reviewer as simply stronger than the puppet master, and the puppet master itself is described as prone to falling apart entirely once its board position goes wrong.

No visible support since launch week

This review found nothing in the game's public patch history beyond a short cluster of five posts, all dated to the week of the 19 December 2024 launch — a Mac beta release and two minor bugfix patches. Nothing has followed since, as far as this review can confirm, despite at least one reviewer expressing direct hope for "more characters, more maps" in future content. Given the specific bug reports above, and the detailed structural criticism of the game's fixed node sequence, it's reasonable to read the current build as close to its final state rather than one still actively evolving.

Final verdict

Tetra Dungeon's central idea — cards as literal Tetris pieces determining combat geometry — is a genuinely clever, well-executed hook that easily justifies its low asking price for a handful of short, satisfying sessions. But the detailed, credible criticism that its roguelike structure is mostly fixed rather than randomised, paired with an apparent lack of any post-launch content updates, means this is a small, complete experience rather than the deep, replayable card-roguelike its genre tags might suggest. Buy it for a clever afternoon or two with Tetris and Slay the Spire in mind, and don't expect it to hold up to the scrutiny genre veterans will inevitably apply.