Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Unity Netcode for GameObjects skill for multiplayer networking, RPCs, state synchronization, and server-authoritative gameplay.
.claude/skills/a5c-ai-unity-netcode/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-19 | ✗→✓ | ▲ Improved | 89% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 10% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 4% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 4% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 67% | 0% |
Multiplayer networking using Unity Netcode for GameObjects.
This skill provides capabilities for implementing multiplayer games using Unity's Netcode for GameObjects, including state synchronization, RPCs, and network topology configuration.
csharppublic class PlayerNetworkBehaviour : NetworkBehaviour { private NetworkVariable<int> health = new NetworkVariable<int>(100); public override void OnNetworkSpawn() { if (IsOwner) { // Initialize owned player } } [ServerRpc] public void TakeDamageServerRpc(int damage) { health.Value -= damage; if (health.Value <= 0) { DieClientRpc(); } } [ClientRpc] private void DieClientRpc() { // Play death effects on all clients } }
csharp// Configure in NetworkManager component networkManager.ConnectionApprovalCallback = ApprovalCheck; void ApprovalCheck(NetworkManager.ConnectionApprovalRequest request, NetworkManager.ConnectionApprovalResponse response) { response.Approved = ValidatePlayer(request.Payload); response.CreatePlayerObject = true; }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 10,033 | 13,639 | +36% | 1 | 1 | 0% | 1,945 | 2,137 | +10% | 0 | 0 | — |
case-02 | pass→pass | 16,576 | 14,120 | -15% | 1 | 1 | 0% | 2,126 | 2,215 | +4% | 0 | 0 | — |
case-03 | pass→pass | 16,187 | 13,973 | -14% | 1 | 1 | 0% | 2,182 | 2,278 | +4% | 0 | 0 | — |
case-04 | pass→pass | 4,360 | 9,298 | +113% | 1 | 1 | 0% | 733 | 1,221 | +67% | 0 | 0 | — |
case-05 | pass→pass | 6,835 | 10,218 | +49% | 1 | 1 | 0% | 1,265 | 1,427 | +13% | 0 | 0 | — |
case-06 | pass→pass | 11,155 | 16,916 | +52% | 1 | 1 | 0% | 2,057 | 2,218 | +8% | 0 | 0 | — |
case-07 | pass→pass | 11,428 | 9,615 | -16% | 1 | 1 | 0% | 885 | 1,254 | +42% | 0 | 0 | — |
case-08 | pass→pass | 4,552 | 8,909 | +96% | 1 | 1 | 0% | 726 | 1,136 | +56% | 0 | 0 | — |
case-09 | pass→pass | 17,210 | 15,360 | -11% | 1 | 1 | 0% | 2,171 | 2,041 | -6% | 0 | 0 | — |
case-10 | pass→pass | 16,489 | 13,866 | -16% | 1 | 1 | 0% | 2,096 | 2,102 | +0% | 0 | 0 | — |
case-11 | pass→pass | 17,040 | 18,625 | +9% | 1 | 1 | 0% | 2,296 | 3,194 | +39% | 0 | 0 | — |
case-12 | pass→pass | 11,320 | 15,390 | +36% | 1 | 1 | 0% | 2,116 | 2,605 | +23% | 0 | 0 | — |
case-13 | pass→pass | 10,868 | 10,166 | -6% | 1 | 1 | 0% | 1,053 | 1,166 | +11% | 0 | 0 | — |
case-14 | pass→pass | 16,704 | 15,417 | -8% | 1 | 1 | 0% | 1,384 | 2,005 | +45% | 0 | 0 | — |
case-15 | pass→pass | 56,299 | 20,860 | -63% | 1 | 1 | 0% | 2,917 | 2,585 | -11% | 0 | 0 | — |
case-16 | pass→pass | 10,579 | 14,252 | +35% | 1 | 1 | 0% | 1,637 | 1,922 | +17% | 0 | 0 | — |
case-17 | pass→pass | 8,725 | 8,374 | -4% | 1 | 1 | 0% | 500 | 908 | +82% | 0 | 0 | — |
case-18 | pass→pass | 8,948 | 2,852 | -68% | 1 | 1 | 0% | 557 | 910 | +63% | 0 | 0 | — |
case-19 | fail→pass | 10,654 | 10,055 | -6% | 1 | 1 | 0% | 749 | 1,418 | +89% | 0 | 0 | — |
case-20 | pass→pass | 23,942 | 20,127 | -16% | 1 | 1 | 0% | 3,120 | 3,173 | +2% | 0 | 0 | — |
case-21 | pass→pass | 5,831 | 11,061 | +90% | 1 | 1 | 0% | 938 | 1,439 | +53% | 0 | 0 | — |
case-22 | pass→pass | 5,952 | 1,533 | -74% | 1 | 1 | 0% | 858 | 706 | -18% | 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. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 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.