Three people. 40+ hand-drawn enemies. One question we kept asking.

Every enemy in *Interstellar Sentinel* was drawn by hand. Not generated, not tiled, not outsourced. Drawn. That decision cost us significant time. It also made the game look the way it looks. Here's how we actually built this.

We're three people. The scope of this project — 8 levels, 40+ enemies, 19 weapon sets, 58 minutes of original music — is not normal for a team this size. Every system in our process was designed around keeping the vision intact despite the constraints.

The art process starts with the world, not the mechanics. Before weapons are tuned. Before bullet patterns are designed. Before level sequences are set. Because in a shmup where you can't stop and read the story, the art *is* the story. The design decisions in concept art directly determine what the player feels.

The iteration question was always: *does this belong in this world?* Not "is this well-drawn" or "does this fit the genre" — does it feel like it belongs *here*. If it read as a game element instead of a world element, we pushed it. That filter is why 40+ enemies feel like they're from the same place.

Chris Huelsbeck didn't score a game. He scored scenes. He saw the art and responded to it. We didn't give him a brief — we gave him concept art. The result is a score that moves when the world moves, not when the gameplay mechanics change.

That's not how most games are made. Most game scores are written to gameplay loops. This one was written in conversation with the art. Which is why when you're in a level, the music and the environment feel like they know each other.

You can trace this through the art book. Sketches to structured passes to shipped designs. The through-line is always: does this belong here?

*Thursday: the full art book — what's in it, how to get it, and everything we've been building toward for the past four weeks.*