Choose a contract.
Open the doors, build audience trust, or protect the crew. Each contract has different targets. Hit every target at the end of five calls to fulfill it.
Pressure mode: two forecast scans.
Read the incoming message and select a response. Exact consequences remain concealed until you commit, unless you spend a scan on that response. Each of your two scans reveals one response; switching to another does not reveal it. You can revisit a scanned response without spending again.
Sandbox: see the possibilities.
Every response previews its exact effect. Explore the model at your own pace with the same contracts, scenes, and endings.
One rewind. Then another worldline.
You can undo the most recent committed call once per run. Used scans are not refunded. After the ending, branch from any decision to explore a fresh continuation. A branch starts a new run with fresh tools.
Your crew remembers.
Incoming messages react to prior decisions and the state of the plan. These narrative reactions do not add hidden costs. Consequences, targets, and the pressure indicator come from authored rules.
Time follows the dependencies.
The plan starts at day zero. Tasks can overlap when their dependencies permit. The last task’s finish is the projected opening day. All five calls revise this forecast; the game does not simulate completed work or an elapsed real calendar.
The world is fictional.
You start with $100,000, $72,000 committed, team energy 84, and trust 78. Energy, trust, and pressure are illustrative gauges. The model omits resource leveling and empirical probabilities. It cannot predict an actual project.
Make it comfortable.
Sound starts off. Motion and spatial depth can be switched off. All scenes can be continued immediately; Escape dismisses a scene and leaves its consequence available to read. Your run archive is stored only in this browser, and can be cleared.