Hold The Horde — Expedition Proposal

What this is. One representative sortie from the ship into the station and back, written to show that a clear immediate goal, ordinary horde pressure, a discovery that changes understanding and a purposeful return can be one trip rather than four activities. It is a comparison scenario, not an opening mission and not settled design. Requirements come from the Owner Brief; everything else is a proposal, and revival and the ending stay open. Build rules: Build Comparison. Scope: Design Plan.

Leaving the ship

The colony wants a compressor head: station gear that packs a canister denser than the ship can alone. The ship makes unlimited oxygen (owner requirement); what it cannot make is reach. That is why a successful trip pays out in exploration rather than damage.

Yeva Kolb — a station pressure-tech the player rescued on an earlier trip — names the workshop and one condition: the head has to arrive uncracked. That is the entire immediate goal, and it is checkable at any moment.

The loadout is the first real decision. Equipment competes with salvage for space (owner requirement), and under our working assumption spare canisters compete with both (proposal, not canon — see Dependencies). The head is bulky: reserving room for it means going in lighter than the player would like, and going in heavy means a second trip or an abandoned objective. None of it is hidden — carried supply, remaining time and free space stay readable at a glance. Mystery belongs in the station, never in the interface.

The way in

The route runs down a service spine. The horde here is ordinary — no set piece — and it does what ordinary hordes should do: it forces route commitment. A chokepoint can be fought, sealed or avoided, and each option spends something different: equipment, time and therefore oxygen, or distance.

The player passes things without stopping. A mark cut into a crate. Another beside a pressure door two junctions later — the same unfamiliar script, but hand-painted and irregular where the first was exact. The difference is plainly visible and means nothing yet. Both are scenery, and the player is right to treat them that way.

The rack

The head sits in a powered rack whose state is legible: it holds until unpowered, and unpowering it is a deliberate act the player has to protect. Retrieval is a commitment, not a pickup, and finding the head does not complete the objective; only arriving with it does.

The shutdown is also the first thing the player watches a mark do. It travels visibly through the rack, runs through one cut-marked part and stops dead at another marked differently. Nothing is named. The player has simply seen two marks behave differently, once. This depends on a rule Systems owns and has not yet confirmed — see Dependencies.

What the player notices

The service bay behind the workshop finishes the job the rack started. There the player watched two marks behave differently, once. Here the same shapes recur across components that plainly do different jobs, each staying consistent within its group. One observation becomes a pattern, and a player who compares them can conclude, unprompted, that the marks name families — relay, sink, isolator — rather than individual parts. Evidence first, generalisation second, and the inference is never handed over.

Observed, not unlocked. Every mark was physically visible the first time the player walked past it; nothing is revealed here. No solve button, no survivor who translates. The notebook preserves a facsimile and location of a mark the player chooses to record and does no more. Knowing three families is not knowing a language.
Legible. A player who does not care can take the head and leave unpenalised; a player who looks gets the discovery.
Paid for in oxygen. Standing in that bay is time, and time is supply. This does not dissolve the brief's tension between pressure and curiosity. The intent is that the payment feels worth making; whether it does is exactly the kind of thing writing cannot settle.

The way out

The return trip is the point of the whole design. The corridor is the same corridor. The player is not.

In hand the head leaves only basic defence: the build they spent capacity on is unusable while they are holding it. Stowing it returns the build through a deliberate, exposed transfer, and it still occupies full space. In-hand carriage is brief by design — once stowing is safe the player takes it, and no threat is invented to keep them holding it. The honest recurring decision homeward is stowed against dropped, and dropping frees space and handling at once while leaving a fragile objective on the floor behind them.

The marks they walked past have also stopped being scenery. The crate's cut mark reads as a sink now, making it the trip's last capacity question: make room, or leave it knowing what they are leaving.

The door reads differently again. Its paint is irregular where the cutting was exact — a copy, made by a human hand from a shape the crew did not understand and used as a label of their own. It identifies nothing mechanical and does not pretend to. What it says is that someone deliberately sealed that section, and wanted others to know.

Home

The return is purposeful three times over. The head is installed and the colony's canisters reach further. The knowledge outlives the trip. And Yeva can recognise the paint — but only if she is shown it.

That is the proposed channel, and it earns its keep: showing a facsimile to a survivor is what the notebook is for, and it is the reason to record an irregular scrawl that means nothing at the time. Survivors read handwriting, not language. The consequence is therefore missable rather than automatic — a deliberate cost, since a player who recorded nothing gets no scene — but missable is not lost: the mark stays on the door and Yeva reacts whenever she is shown it.

Shown it, she says it is Dorn's — her work partner, whom she has not named before now. That is the human consequence, and deliberately not a resolution. She does not know whether he sealed the section to keep something out or to keep himself in; neither does the player, and she wants to know. Her contribution is human context — whose hand, which crew practice — while the pattern-finding stayed with the player. She is a person, not an upgrade dispenser.

What the player can now do differently

A named place to go back to. The sealed door is an objective the player chose, not one handed out.
A reading that travels. Cut marks recur station-wide, so the families change which components are worth carry space later; painted copies of the same shapes carry crew context and no mechanical claim. Two kinds of information in one script, told apart by how the mark was made. What the families feed into is Build Comparison's to state, and it rests on the rack observation below.
More reach. The installed head reopens places the player had to turn back from.

Provisional mechanics and dependencies

None of this is owner canon. The scene survives if these change.

The rack observation. That unpowering visibly runs through one marked part and stops at another is proposed here and not yet confirmed by Systems. It is the only observed evidence the family inference rests on. If Systems declines it, the honest correction is to keep the discovery as understanding plus the chosen door and drop the mechanical reading — not to assert a behaviour the player never sees.
Carry rules. In hand, stowed and dropped come from Build Comparison; no magnitude is asserted. Two points are Systems' and this scene follows whatever they set: whether the head occupies capacity while held, which decides whether the loadout paragraph's "going in heavy" cost is real, and confirmation that in-hand carriage is meant to be transient.
Shared carry interface. Equipment versus salvage is an owner requirement; canisters competing for the same space is a working assumption. If oxygen gets separate capacity the loadout decision weakens but the scene holds.
Notebook. Facsimiles and locations of marks the player records, and nothing else. Showing one to a survivor is the proposed use that makes recording worth doing.
Findability across trips. The sealed door must stay locatable later or the human consequence leads nowhere. That needs discovered locations to persist — weaker than a fixed map, which stays open.
Death losses. Undecided, and deliberately not settled here, though what happens to a carried head on death changes the return leg's whole risk character.
Salvage magnitude. How much reach the head adds, and whether it is unique or one of a class, are not mine to set.

What this does not establish

That this is fun; that oxygen pressure tunes to a level rewarding care rather than punishing it; or that horde combat and unhurried observation coexist in the same minute of play. Those are the real risks and a written scene cannot settle them. One narrower claim holds: this discovery reuses signage, corridors and a mechanism that already exist, so its cost is authored meaning rather than new geometry — a reason to prefer this kind of discovery, not evidence about the 30 / 90 / 100–200 hour ambitions, which remain unvalidated.