How to Play

What is We-Go?

CORP//LLM uses the We-Go system: All players plan their actions simultaneously without knowing what others are planning. Once all players are ready, the engine resolves all commands simultaneously. There is no turn order — everything happens “at once”. This means: you can plan an attack on a building while your opponent is simultaneously evacuating that exact building. Whoever thinks further ahead wins.

The Daily Cycle

The daily rhythm is player-driven — no fixed schedule:

Phase	Trigger	What happens
Planning Phase	after the last tick	Issue commands to your runners, buy buildings, set conditions
READY	each player individually	Click “READY” in the dashboard when you’re done planning
Simulation (Tick)	all players ready	The engine resolves all actions simultaneously (1,440 minute-ticks)
Results	after simulation	Morning Gazette published, new day begins

Fallback: If not all players are ready within 24 hours, the day is automatically resolved.

One tick equals one day in the game world. If everyone plans quickly, multiple days can occur per real-time day.

Session Modes

Not every game runs for the same length. When creating a session, the host chooses a mode:

Mode	Cycles (Days)	Sectors / Player	Best for
Quick	20	3	Fast rounds, ideal for new players
Standard	30	5	The full experience
Epic	45	6	Long-term strategies, large maps

The actual city size scales dynamically with player count: citySize = max(4, effectivePlayers) × sectorsPerPlayer[mode]. Standard with 6 players = 30 sectors, Epic with 8 players = 48 sectors. Engine hard cap: 20 players.

Session Lifecycle

Each session goes through three phases:

Hub — After login (Google, GitHub, or Discord), you land on the Hub with three tabs: Sessions (create/join), Profile (account linking, stats), and Settings. In Settings you can customize your player character — model, skin tone, pose and accessories. A 3D profile picture (head portrait) is automatically rendered and saved from your character. When creating a session, you can configure bot opponents (temperament + difficulty) and optionally set a password for private games. At the bottom of the Sessions tab, beta testers see two widgets: Roadmap and Bugtracker (open items from GitHub project boards — completed items are hidden). Below that, the version is shown as client vX.Y.Z (hash) · server vA.B.C (hash) plus test coverage for both repos.
Lobby — After creating a session, you’re in the lobby. Other players can join (password-protected sessions require the password). As the creator, you can click “Start Game” at any time — even solo. Alternatively, the game starts automatically when all players are ready or the deadline hour is reached. Empty slots are filled with AI bots.
Running — The session is live. Daily planning and simulation cycles alternate.
Finished — The final tick is done. Final rankings, statistics, and replay available.

After the Tick: Morning Briefing

Every morning after the simulation, the Morning Briefing Overlay opens automatically. At a glance you’ll see:

Finances — Revenue, expenses, balance, laundering summary
Runner Status — Who is injured, arrested, killed, or deserted?
Gazette Highlights — Top headlines from the Morning Gazette: raid in Sector 14, police crackdown at the docks, mysterious warehouse fire…

Then close the overlay and begin planning for the next day.

Timeline Playback

Want to know exactly what happened last night? Timeline Playback shows you the entire tick in chronological order:

Who moved where?
Which actions were executed — and with what result?
Where did player clashes occur?

Use the Timeline for forensic analysis: why was your runner caught? Who set fire to the warehouse? The answers are in the timeline.

Tip: Plan with Fallbacks

You can’t predict everything. That’s why CORP//LLM has a condition system: instead of just planning one action, you can build IF/THEN/ELSE chains. For example:

If my runner has more than 50 HP, then RAID the warehouse in Sector 12. Otherwise: FLEE_TO_SAFEHOUSE.

Your runners automatically respond to situations you couldn’t foresee. Learn more on the Actions & Planning page.

AI Advisor

Using ChatGPT, Claude, or another LLM for strategy advice? CORP//LLM has a built-in AI Advisor (screen #llm). It prepares your complete game state as structured JSON — runners, buildings, economy, diplomacy, gazette news — and generates a system prompt that explains the game rules to the LLM.

Workflow:

Copy the system prompt and paste it into your LLM
Copy the game state export (or full export) and send it along
The LLM analyzes your situation and suggests an optimized plan
Paste the plan JSON into the import area, validate, and submit directly

The plan is sent straight to the server — no manual copy-paste needed.

Export blocks:

Block	Contents
Runner Overview	Name, stats (HP/Heat/Loy/Stress), role, team
Buildings & Sectors	Owned buildings + sector info
Economy	Cash, net worth, federal heat, daily balance
Diplomacy	NAPs, rivals, intel
Current Plan	Submitted actions for the current cycle
Gazette / News	Latest headlines and events
Sector Buildings (FoW)	All buildings in scouted sectors (world buildings)
Full Export	All data incl. leaderboard, gazette, world buildings, available actions

The system prompt automatically includes: runner stats, scouted sectors, current plan, and latest gazette headlines.

Fog of War

At the start of the game you can only see your home sector and its adjacent sectors. All other areas of the city are hidden.

Discovering new sectors:

RECON_SECTOR — Immediate action (no plan needed). Cost: $1,000 clean cash.
Reveals a random adjacent undiscovered sector.
Additionally, neighbors of the new sector also become visible.
Once revealed = permanently visible (no re-fog).

What is restricted?

The city map only shows scouted sectors with buildings and details.
The AI Advisor only exports buildings in scouted sectors.
All fixture data (sectors, heat, ownership) is server-side filtered — no client hack possible.