Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Unity Visual Effect Graph skill for GPU particle systems, procedural effects, and high-performance visual effects.
.claude/skills/a5c-ai-unity-vfx-graph/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 14% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 11% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 5% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 21% | 0% |
Visual Effect Graph development for GPU-accelerated particle systems in Unity.
This skill provides capabilities for creating high-performance visual effects using Unity's Visual Effect Graph, leveraging GPU compute for millions of particles.
csharppublic class VFXController : MonoBehaviour { [SerializeField] private VisualEffect vfx; void Start() { // Set properties vfx.SetFloat("SpawnRate", 100f); vfx.SetVector3("EmitterPosition", transform.position); } public void TriggerBurst() { // Send event vfx.SendEvent("OnBurst"); } }
csharp// Expose VFX properties via C# bindings vfx.SetInt("ParticleCount", 1000); vfx.SetGradient("ColorOverLife", gradient); vfx.SetTexture("ParticleTexture", texture);
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 10,192 | 16,931 | +66% | 1 | 1 | 0% | 1,810 | 2,062 | +14% | 0 | 0 | — |
case-02 | pass→pass | 7,686 | 8,001 | +4% | 1 | 1 | 0% | 1,462 | 1,627 | +11% | 0 | 0 | — |
case-03 | pass→pass | 7,827 | 5,824 | -26% | 1 | 1 | 0% | 1,431 | 1,499 | +5% | 0 | 0 | — |
case-04 | pass→pass | 4,598 | 4,172 | -9% | 1 | 1 | 0% | 890 | 1,076 | +21% | 0 | 0 | — |
case-05 | pass→pass | 5,506 | 3,858 | -30% | 1 | 1 | 0% | 978 | 1,015 | +4% | 0 | 0 | — |
case-06 | pass→pass | 8,426 | 5,489 | -35% | 1 | 1 | 0% | 1,480 | 1,454 | -2% | 0 | 0 | — |
case-07 | pass→pass | 16,367 | 8,929 | -45% | 1 | 1 | 0% | 2,384 | 1,942 | -19% | 0 | 0 | — |
case-08 | pass→pass | 14,204 | 9,481 | -33% | 1 | 1 | 0% | 1,979 | 2,007 | +1% | 0 | 0 | — |
case-09 | pass→pass | 4,851 | 2,095 | -57% | 1 | 1 | 0% | 759 | 779 | +3% | 0 | 0 | — |
case-10 | pass→pass | 4,361 | 2,917 | -33% | 1 | 1 | 0% | 711 | 824 | +16% | 0 | 0 | — |
case-11 | fail→pass | 8,791 | 1,797 | -80% | 1 | 1 | 0% | 1,570 | 656 | -58% | 0 | 0 | — |
case-12 | pass→pass | 5,245 | 7,216 | +38% | 1 | 1 | 0% | 907 | 1,686 | +86% | 0 | 0 | — |
case-13 | pass→pass | 5,931 | 6,417 | +8% | 1 | 1 | 0% | 936 | 1,449 | +55% | 0 | 0 | — |
case-14 | pass→pass | 14,588 | 13,641 | -6% | 1 | 1 | 0% | 2,275 | 2,620 | +15% | 0 | 0 | — |
case-15 | pass→pass | 10,689 | 9,760 | -9% | 1 | 1 | 0% | 1,455 | 2,086 | +43% | 0 | 0 | — |
case-16 | pass→pass | 18,536 | 13,582 | -27% | 1 | 1 | 0% | 2,445 | 2,512 | +3% | 0 | 0 | — |
case-17 | pass→pass | 17,440 | 13,363 | -23% | 1 | 1 | 0% | 2,466 | 2,689 | +9% | 0 | 0 | — |
case-18 | pass→pass | 18,071 | 13,868 | -23% | 1 | 1 | 0% | 2,649 | 2,939 | +11% | 0 | 0 | — |
case-19 | pass→pass | 9,209 | 6,789 | -26% | 1 | 1 | 0% | 1,489 | 1,501 | +1% | 0 | 0 | — |
case-20 | pass→pass | 11,080 | 11,315 | +2% | 1 | 1 | 0% | 2,045 | 2,644 | +29% | 0 | 0 | — |
case-21 | pass→pass | 9,942 | 7,251 | -27% | 1 | 1 | 0% | 1,799 | 1,858 | +3% | 0 | 0 | — |
case-22 | pass→pass | 10,434 | 11,269 | +8% | 1 | 1 | 0% | 2,004 | 2,817 | +41% | 0 | 0 | — |
case-23 | pass→pass | 12,878 | 9,828 | -24% | 1 | 1 | 0% | 2,449 | 2,337 | -5% | 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. 23 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 23 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.