Gunhead

Alientrap has always been the studio that does the interesting thing instead of the safe thing. Capsized was a physics-driven jetpack platformer when nobody wanted one. Apotheon rendered an entire action game in Greek black-figure pottery, and it looked like nothing else on the market before or since. Cryptark, in 2017, took the roguelite — then busy eating the entire indie space — and did something almost perverse with it: it made the shooting secondary. You flew a salvage mech into a derelict alien hulk, and the actual game was a systems puzzle. Which subsystem do you kill first? The one feeding the shields, or the one spawning repair drones that will undo your work while you’re busy elsewhere?

Gunhead is that game rebuilt in the first person, and the translation is more faithful than you might expect. You’re still a privateer with a cannon where your head should be. You’re still boarding procedurally generated space hulks stitched out of bone and steel. The objective is still the System Core at the vessel’s heart, and the path there is still gated by a lattice of interlocking security systems you can choose to dismantle, ignore, or blunder into. What’s new is verticality: a jetpack, a full three axes of movement, and the messy, fast-twitch pressure of having threats above and behind you rather than politely arranged on a 2D plane.

Positioning-wise, this is a niche product and it knows it. It is not competing with Doom Eternal or Returnal. It’s competing for the specific player who plays Hitman for the planning, not the killing — the one who wants a shooter that rewards a pre-mission read of the blueprint. That audience exists, it’s just small, and Alientrap made no real effort to court anyone outside it. The result is a game that reads less like “flawed” and more like “aimed at the wrong half of the people who bought it” — which is exactly why it lands as a mixed recommendation rather than an easy one.

• Subsystem-Driven Objectives: Every hulk’s defenses — shields, alarms, repair drones, turret networks — are physically located on discrete, destructible subsystems. Killing the right one collapses half the ship’s resistance. Killing the wrong one wastes a minute you can’t get back.
• The Payout Clock: Your contract fee bleeds out in real time from the moment you board. Speed and thoroughness are directly opposed, and the game never lets you have both.
• Four-Slot Loadout from 50+ Weapons: Plus mech chassis and support systems purchased between contracts. Builds are divergent, not cosmetic re-skins.
• Jetpack Traversal: Fast, free, and vertical. The single biggest thing 3D adds over Cryptark.
• Persistent Non-Linear Campaign: Runs aren’t isolated attempts. Funds, gear, and route decisions carry forward, so a bad contract has consequences three missions later.

The Gameplay Loop

Most roguelite shooters ask one question per room: can you kill this? Gunhead asks a better one: what is this room connected to?

Here’s how a contract plays. You get a manifest before you board — a list of the ship’s active systems and a rough map. This is the real game, and it happens before you fire a shot. You see the hulk is running a shield generator and an alarm system; you know the alarm will keep summoning reinforcements, and you know the shields will make the Core itself immune until they’re down. You have finite health, a decaying payout, and no ability to be everywhere at once. So you pick an order, commit, and find out how wrong you were.

That commitment is what makes the loop work. Because subsystems are spatially real — actual rooms in actual corridors — your plan is also a route, and the route is where the improvisation lives. Three rooms into a clean approach, a repair drone swarm can undo a door you already cut open, and suddenly the efficient path is gone and the math is about whether to backtrack or eat the damage. The pressure is cognitive, not mechanical, and that’s a rarer feeling than it should be in this genre.

The Economy Is the Difficulty Curve

The payout clock is the sharpest piece of design here. Time is money, literally: linger to strip a ship clean and you’ll finish with a completed objective and a fee not worth the ammunition. Rush it and you’ll arrive at the Core underequipped, with the alarm still screaming and a drone factory still running.

Because the campaign is persistent, those choices compound. A greedy, sloppy contract early doesn’t just cost you that mission — it under-funds your next chassis, which makes the following contract harder, which makes you greedier. The game generates its own difficulty curve out of your own bad decisions. That’s elegant, and it’s the thing Gunhead does that almost nothing else in the first-person space attempts.

Combat and Readability

The moment-to-moment shooting is competent rather than exceptional. Weapons have weight and clear roles — crowd clear, single-target, structural damage — and the four-slot constraint forces real trade-offs, since no build covers every situation. Enemies are readable by silhouette, which matters enormously when you’re inverted under a ceiling with a jetpack running hot.

The weak link is spatial legibility. Procedural interiors are hard, and Alientrap’s are functional but rarely memorable: corridors, junctions, bulkheads. In a game whose central skill is route planning, “where am I” should never be a live question, and too often it is. The minimap does honest work, but the design demands more consultation with it than it should. A stronger hand with landmarking — distinct architecture per subsystem type, say — would have paid for itself ten times over.

Where It Wears Thin

Repetition is the game’s most consistent weakness, and the criticism is fair. The generator has a limited vocabulary. Well before the campaign ends you’ve seen the shapes, and once the layouts stop surprising you, the strategic layer has to carry the entire experience alone. It’s strong enough to do that for a while, but not indefinitely. The narrative framing — privateer, contracts, salvage — is set dressing, not a reason to continue. There’s no escalating fiction pulling you forward the way Hades or Returnal pull.

Taken in short bursts of two or three contracts, this is a strong game. As a binge, the seams show fast — a real problem for a persistent campaign built to be lived in.

This is the part that stings, because Alientrap of all studios should have known better.

Gunhead is flat cel-shaded sci-fi. Clean lines, hard outlines, muted palette. It is perfectly legible and totally forgettable — the visual equivalent of a competent stock font. From the team that built an entire game inside an ancient Greek amphora, “inoffensive shaded polygons” is a genuine disappointment. The bone-and-steel enemy design is the one place with real personality, and it deserved a world built to match.

The design also leaves little margin for technical error: with a payout clock ticking and enemies attacking from three axes, any graphical hiccup at the wrong moment costs money, so the game’s heavier encounters put real demands on the hardware and on Alientrap’s own engine work. The UI is the opposite of the art — dense, information-forward, and well-organized. The pre-mission manifest screen does a lot of heavy lifting with minimal friction.

Steam, mouse and keyboard, and that’s the right and arguably the only way to play it.

The combination of jetpack verticality and precision targeting of small subsystem components is a mouse’s job. Controller support exists and functions, but the aim assist can’t fully compensate for the fine-grained, fast-panning targeting the game asks for once a room goes loud. Deck-style handheld play is viable in the mechanical sense and miserable in the practical one — the text-dense mission briefings and loadout screens were designed for a monitor at desk distance.

The Steam context also matters for a different reason: this is a game whose store-page score is actively misleading. It averages people who loved a strategy game with people who bought an action game. Read the negative reviews before you buy; they’ll tell you very quickly whether you’re the second group.

Gunhead is a good game that could have been a significant one. The core idea — a shooter where the hardest problem is deciding what to destroy first — is sharp, well-executed, and still under-explored by the rest of the industry. The persistent campaign and the payout clock turn what could have been a repetitive score-chase into something with actual consequence.

What’s missing is ambition everywhere except the systems. The art direction is safe, the fiction is filler, and the procedural generator is too small a box for the campaign it has to fill. Alientrap made the sequel work; they didn’t make it soar. For the specific player it’s aimed at, that’s a qualified recommendation. For most people, the mixed reception is telling you something true — and we rate it accordingly.