Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Unreal Engine Gameplay Ability System (GAS) skill for attributes, abilities, and gameplay effects.
.claude/skills/a5c-ai-unreal-gamesframework/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-16 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-01 | ✓→✓ | = Same ✓ | -17% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 0% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 14% | 0% |
| case-04 | ✓→✓ | = Same ✓ | -5% | 0% |
Gameplay Ability System for Unreal Engine.
This skill provides capabilities for implementing complex gameplay systems using Unreal's Gameplay Ability System (GAS).
cppUCLASS() class UMyAttributeSet : public UAttributeSet { GENERATED_BODY() public: UPROPERTY(BlueprintReadOnly, ReplicatedUsing=OnRep_Health) FGameplayAttributeData Health; ATTRIBUTE_ACCESSORS(UMyAttributeSet, Health) UPROPERTY(BlueprintReadOnly, ReplicatedUsing=OnRep_MaxHealth) FGameplayAttributeData MaxHealth; ATTRIBUTE_ACCESSORS(UMyAttributeSet, MaxHealth) };
cppUCLASS() class UGA_Attack : public UGameplayAbility { GENERATED_BODY() protected: virtual void ActivateAbility( const FGameplayAbilitySpecHandle Handle, const FGameplayAbilityActorInfo* ActorInfo, const FGameplayAbilityActivationInfo ActivationInfo, const FGameplayEventData* TriggerEventData) override; };
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 15,442 | 13,148 | -15% | 1 | 1 | 0% | 2,557 | 2,119 | -17% | 0 | 0 | — |
case-02 | pass→pass | 11,792 | 8,640 | -27% | 1 | 1 | 0% | 1,869 | 1,875 | +0% | 0 | 0 | — |
case-03 | pass→pass | 10,220 | 10,602 | +4% | 1 | 1 | 0% | 1,836 | 2,086 | +14% | 0 | 0 | — |
case-04 | pass→pass | 11,368 | 7,887 | -31% | 1 | 1 | 0% | 1,857 | 1,765 | -5% | 0 | 0 | — |
case-05 | pass→pass | 10,040 | 5,052 | -50% | 1 | 1 | 0% | 2,011 | 1,185 | -41% | 0 | 0 | — |
case-06 | pass→pass | 5,683 | 2,851 | -50% | 1 | 1 | 0% | 813 | 984 | +21% | 0 | 0 | — |
case-07 | pass→pass | 8,171 | 5,796 | -29% | 1 | 1 | 0% | 1,387 | 1,669 | +20% | 0 | 0 | — |
case-08 | pass→pass | 7,148 | 5,111 | -28% | 1 | 1 | 0% | 1,192 | 1,353 | +14% | 0 | 0 | — |
case-09 | pass→pass | 5,148 | 4,506 | -12% | 1 | 1 | 0% | 1,059 | 1,137 | +7% | 0 | 0 | — |
case-10 | pass→pass | 3,472 | 3,871 | +11% | 1 | 1 | 0% | 640 | 1,193 | +86% | 0 | 0 | — |
case-11 | pass→pass | 14,313 | 2,843 | -80% | 1 | 1 | 0% | 2,150 | 891 | -59% | 0 | 0 | — |
case-12 | pass→pass | 7,396 | 6,581 | -11% | 1 | 1 | 0% | 1,228 | 1,577 | +28% | 0 | 0 | — |
case-13 | pass→pass | 17,711 | 16,016 | -10% | 1 | 1 | 0% | 2,533 | 3,030 | +20% | 0 | 0 | — |
case-14 | pass→pass | 18,843 | 23,655 | +26% | 1 | 1 | 0% | 2,954 | 3,713 | +26% | 0 | 0 | — |
case-15 | pass→pass | 12,614 | 6,963 | -45% | 1 | 1 | 0% | 1,974 | 1,441 | -27% | 0 | 0 | — |
case-16 | fail→pass | 13,725 | 10,711 | -22% | 1 | 1 | 0% | 1,896 | 2,231 | +18% | 0 | 0 | — |
case-17 | pass→pass | 8,967 | 10,498 | +17% | 1 | 1 | 0% | 1,623 | 2,027 | +25% | 0 | 0 | — |
case-18 | pass→pass | 19,432 | 18,593 | -4% | 1 | 1 | 0% | 3,245 | 3,829 | +18% | 0 | 0 | — |
case-19 | pass→pass | 12,810 | 12,319 | -4% | 1 | 1 | 0% | 2,244 | 2,447 | +9% | 0 | 0 | — |
case-20 | pass→pass | 14,054 | 15,608 | +11% | 1 | 1 | 0% | 2,278 | 3,761 | +65% | 0 | 0 | — |
case-21 | pass→pass | 14,602 | 20,021 | +37% | 1 | 1 | 0% | 2,903 | 3,790 | +31% | 0 | 0 | — |
case-22 | pass→pass | 17,483 | 16,195 | -7% | 1 | 1 | 0% | 3,011 | 3,573 | +19% | 0 | 0 | — |
case-23 | pass→pass | 20,769 | 23,873 | +15% | 1 | 1 | 0% | 4,155 | 5,462 | +31% | 0 | 0 | — |
case-24 | pass→pass | 9,973 | 8,638 | -13% | 1 | 1 | 0% | 2,058 | 2,161 | +5% | 0 | 0 | — |
case-25 | pass→pass | 17,277 | 15,185 | -12% | 1 | 1 | 0% | 3,752 | 3,756 | +0% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 25 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.