Skip to content

Brain Designer

The Brain Designer occupies the center of the GOAP Hub. It provides two complementary graph views: the Planner Graph (top) and the Neural Circuit (bottom).

Planner Graph

The Planner Graph is a node-based view of the currently selected behavioral state. It shows all actions, their connections, and how the planner sees the state.

Layout

Nodes are organized into columns from left to right:

Column Contains
Beliefs All beliefs used as preconditions or effects in this state
Actions Action nodes showing strategy, cost, and post-processes
Goals Goal nodes showing priority and desired state conditions
Transitions State transition targets (if any goal has a TransitionToStateId)

Edges

  • Precondition edges connect beliefs to actions — these beliefs must be true for the planner to consider the action.
  • Effect edges connect actions to beliefs — these are the beliefs the action promises to make true.
  • Goal edges connect beliefs to goals — these are the desired state conditions the planner works backward from.

Reading the Graph

The planner works right to left: it starts from a goal's desired conditions, then searches for actions whose effects satisfy those conditions, checking that preconditions are met or can be achieved by earlier actions.

Node Selection

Click any node to select it. The Inspector panel on the right updates to show that node's properties for editing. You can:

  • Edit action costs, strategies, and settings
  • Configure preconditions and effects
  • Adjust goal priorities and validity conditions
  • Set up post-process chains

Neural Circuit

The Neural Circuit is the brain-level state map. It shows all behavioral states and the transitions between them.

Layout

  • Each state appears as a node
  • Transition arrows connect states, labeled with their belief condition
  • The default state is visually distinguished
  • The current state (in Debug mode) is highlighted

Click a state node in the Neural Circuit to switch the Planner Graph above to that state's view. This is a quick way to navigate between states without using the Library panel or state tabs.

Transitions

Each arrow represents a GoapStateTransition:

  • Source — the state the transition fires from
  • Target — the state the agent moves to
  • Condition — the belief that triggers the transition
  • Target Value — whether the belief must be true or false

Note

Transitions are checked on a timer (default 0.1s). When a transition fires, it interrupts the current plan, calls OnStop on the active strategy, and requests a new plan in the target state.

Inspector Panel

The right side of the GOAP Hub is a context-sensitive property editor. It shows different content depending on what's selected:

Selected Inspector Shows
Action node Strategy SO, settings, cost, cost modifiers, preconditions, effects, post-process chain
Goal node Priority, priority modifiers, desired state conditions, validity conditions, optional state transition
Belief node Belief SO, description, default settings, override settings type
Sensor Sensor SO, update interval, output mappings (slot → key)
Transition Source/target states, condition belief, target value
Key Key name, type, default value

All changes in the Inspector are live — they serialize immediately to the Brain asset.

Auto-Layout

The Planner Graph automatically arranges nodes into columns. When you add new actions, goals, or beliefs, they are placed in the appropriate column without manual positioning.

Design Mode vs Debug Mode

Design Mode (Edit Time)

  • Full editing capability — create, delete, and modify all assets
  • Planner Graph shows the static structure of each state
  • Neural Circuit shows all states and transitions

Debug Mode (Play Mode)

  • Read-only view with live runtime data overlaid
  • Action nodes show current execution state (running, completed, pending)
  • Goal nodes show current priority values (including score modifier output)
  • The active plan is highlighted step by step
  • See Runtime Debugging for details

What's Next