ue-actor-component-arc…
Use this skill when working with Actor and component design in Unreal Engine. Triggers on:…
Use this skill when working with State Tree, StateTree, UStateTree, state machine, StateTreeTask, StateTreeCondition, StateTreeEvaluator, StateTreeSchema, AI State Tree, Mass StateTree, FStateTreeExecutionContext, or data-driven state logic in Unreal Engine. See
$ npx -y skills add quodsoler/unreal-engine-skills --skill ue-state-trees --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/ue-state-treesContext preview
The summary Claude sees to decide when to auto-load this skill.
Use this skill when working with State Tree, StateTree, UStateTree, state machine, StateTreeTask, StateTreeCondition, StateTreeEvaluator, StateTreeSchema, AI State Tree, Mass StateTree, FStateTreeExecutionContext, or data-driven state logic in Unreal Engine. See
name: ue-state-trees description: "Use this skill when working with State Tree, StateTree, UStateTree, state machine, StateTreeTask, StateTreeCondition, StateTreeEvaluator, StateTreeSchema, AI State Tree, Mass StateTree, FStateTreeExecutionContext, or data-driven state logic in Unreal Engine. See references/state-tree-patterns.md for task/condition/evaluator templates and references/state-tree-mass-integration.md for Mass Entity integration." metadata: version: 1.0.0
You are an expert in Unreal Engine's State Tree system for building flexible, data-driven state machines.
Read `.agents/ue-project-context.md` to determine:
Before implementing, clarify: 1. What is the use case? (AI behavior, game logic, UI state, entity processing) 2. What scale? (single actor with `UStateTreeComponent` vs thousands of Mass entities) 3. How complex? (simple linear FSM vs hierarchical states with linked subtrees) 4. Are there existing behavior trees to migrate from? 5. What external data do tasks need? (actor references, subsystems, world state)
---
A State Tree is a hierarchical finite state machine authored as a `UStateTree` data asset:
UStateTree (UDataAsset) ├── UStateTreeSchema ← defines allowed context/external data ├── States[] ← hierarchical state tree │ ├── Tasks[] ← work performed while state is active │ ├── Transitions[] ← rules for leaving this state │ └── Conditions[] ← gates on transitions ├── Evaluators[] ← global data providers (tick before transitions) └── Parameters ← FInstancedPropertyBag default inputs
**Runtime flow per tick:** 1) Evaluators tick, 2) Transitions checked from active leaf up to root, 3) If transition fires: ExitState on old tasks then EnterState on new, 4) Active tasks tick.
**Key classes:**
| Class | Role | |-------|------| | `UStateTree` | Data asset — call `IsReadyToRun()` before execution | | `FStateTreeExecutionContext` | Per-tick context — constructed each frame, NOT persisted | | `FStateTreeInstanceData` | Persistent runtime state — survives across ticks | | `UStateTreeComponent` | Actor component that manages tree lifecycle | | `EStateTreeRunStatus` | `Running`, `Stopped`, `Succeeded`, `Failed`, `Unset` |
**Build.cs modules**: `StateTreeModule`, `GameplayStateTreeModule`
The execution context is constructed per-tick from persistent instance data:
FStateTreeInstanceData InstanceData; // persists across frames // Each tick: FStateTreeExecutionContext Context(Owner, *StateTree, InstanceData); Context.Tick(DeltaTime);
This separates mutable state (`FStateTreeInstanceData`) from stateless execution logic, making State Trees safe for parallel evaluation in Mass Entity scenarios.
---
Schemas define what context data a State Tree can access, constraining valid tasks and conditions. This prevents authoring errors at edit time rather than runtime.
| Schema | Context Provided | Use Case | |--------|-----------------|----------| | `UStateTreeComponentSchema` | Actor + BrainComponent | General actor logic | | `UStateTreeAIComponentSchema` | Above + `AIControllerClass` | AI behavior | | `UMassStateTreeSchema` | Mass entity context | Mass Entity processing |
`UStateTreeComponentSchema` exposes `ContextActorClass` (`TSubclassOf<AActor>`) so the editor knows which components are available for property binding. `UStateTreeAIComponentSchema` extends it with `AIControllerClass` (`TSubclassOf<AAIController>`).
Subclass `UStateTreeSchema` for project-specific trees:
UCLASS()
class UMyGameSchema : public UStateTreeSchema
{
GENERATED_BODY()
public:
virtual bool IsStructAllowed(const UScriptStruct* InStruct) const override;
virtual bool IsExternalItemAllowed(const UStruct& InStruct) const override;
virtual TConstArrayView<FStateTreeExternalDataDesc> GetContextDataDescs() const override;
#if WITH_EDITOR
virtual bool AllowEvaluators() const override { return true; }
virtual bool AllowMultipleTasks() const override { return true; }
virtual bool AllowGlobalParameters() const override { return true; }
#endif // WITH_EDITOR
};Override `GetContextDataDescs()` to declare context objects (actor refs, subsystems). The editor uses this to validate property bindings.
---
Tasks are the primary work units in a state. They are USTRUCTs (not UObjects), making them lightweight and cache-friendly.
Key virtuals (all `const` — tasks are immutable at runtime):
| Virtual | Returns | Called When | |---------|---------|-------------| | `EnterState(Context, Transition)` | `EStateTreeRunStatus` (default: Running) | State becomes active | | `ExitState(Context, Transition)` | `void` | State is exited | | `Tick(Context, DeltaTime)` | `EStateTreeRunStatus` (default: Running) | Each frame (if `bShouldCallTick`) | | `StateCompleted(Context, Status, CompletedStates)` | `void` | Child state completes (REVERSE order) | | `TriggerTransitions(Context)` | `void` | Only if `bShouldAffectTransitions` |
| Flag | Default | Purpose | |------|---------|---------| | `bShouldStateChangeOnReselect` | `true` | Exit+Enter when transitioning to same state | | `bShouldCallTick` | `true` | Enable per-frame Tick calls | | `bShouldCallTickOnlyOnEvents` | `false` | Tick only when events are pending | | `bShouldCopyBoundPropertiesOnTick` | `true` | Refresh property bindings each tick | | `bShouldAffectTransitions` | `false` | Enable `TriggerTransitions` calls |
Set `bShouldCallTick = false` for fire-an
A collection of 27 AI agent skills for Unreal Engine C++ development. Built for game developers who want AI coding agents to help write correct, production-quality UE5 C++ code.
Use this skill when working with Actor and component design in Unreal Engine. Triggers on:…
Use this skill when implementing AI, AIController, behavior tree, blackboard, AI perception,…
Use this skill when working with Unreal Engine animation: AnimInstance, montage playback,…
Use this skill when working with Unreal Engine async operations, threading, parallel…
Use this skill when working with audio, sound, music, UAudioComponent, PlaySoundAtLocation,…
Use this skill when working with character movement, CharacterMovementComponent, CMC,…