Install any skill in seconds. Free to start, no credit card required.
Get Started Free →PC and console game development principles. Engine selection, platform features, optimization strategies.
.claude/skills/lingxling-pc-games/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-21 | ✗→✓ | ▲ Improved | 155% | 0% |
| case-11 | ✓→✓ | = Same ✓ | -22% | 0% |
| case-17 | ✓→✓ | = Same ✓ | 34% | 0% |
| case-01 | ✓→✓ | = Same ✓ | 2% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 29% | 0% |
> Engine selection and platform-specific principles.
What are you building?
│
├── 2D Game
│ ├── Open source important? → Godot
│ └── Large team/assets? → Unity
│
├── 3D Game
│ ├── AAA visual quality? → Unreal
│ ├── Cross-platform priority? → Unity
│ └── Indie/open source? → Godot 4
│
└── Specific Needs
├── DOTS performance? → Unity
├── Nanite/Lumen? → Unreal
└── Lightweight? → Godot| Factor | Unity 6 | Godot 4 | Unreal 5 | |--------|---------|---------|----------| | 2D | Good | Excellent | Limited | | 3D | Good | Good | Excellent | | Learning | Medium | Easy | Hard | | Cost | Revenue share | Free | 5% after $1M | | Team | Any | Solo-Medium | Medium-Large |
| Feature | Purpose | |---------|---------| | Achievements | Player goals | | Cloud Saves | Cross-device progress | | Leaderboards | Competition | | Workshop | User mods | | Rich Presence | Show in-game status |
| Platform | Certification | |----------|--------------| | PlayStation | TRC compliance | | Xbox | XR compliance | | Nintendo | Lotcheck |
Map ACTIONS, not buttons:
- "confirm" → A (Xbox), Cross (PS), B (Nintendo)
- "cancel" → B (Xbox), Circle (PS), A (Nintendo)| Intensity | Use | |-----------|-----| | Light | UI feedback | | Medium | Impacts | | Heavy | Major events |
| Engine | Tool | |--------|------| | Unity | Profiler Window | | Godot | Debugger → Profiler | | Unreal | Unreal Insights |
| Bottleneck | Solution | |------------|----------| | Draw calls | Batching, atlases | | GC spikes | Object pooling | | Physics | Simpler colliders | | Shaders | LOD shaders |
| ❌ Don't | ✅ Do | |----------|-------| | Choose engine by hype | Choose by project needs | | Ignore platform guidelines | Study certification requirements | | Hardcode input buttons | Abstract to actions | | Skip profiling | Profile early and often |
> Remember: Engine is a tool. Master the principles, then adapt to any engine.
This skill is applicable to execute the workflow or actions described in the overview.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | pass→pass | 21,064 | 10,075 | -52% | 1 | 1 | 0% | 2,892 | 2,249 | -22% | 0 | 0 | — |
case-17 | pass→pass | 23,308 | 24,289 | +4% | 1 | 1 | 0% | 2,924 | 3,931 | +34% | 0 | 0 | — |
case-01 | pass→pass | 32,708 | 41,873 | +28% | 1 | 1 | 0% | 4,624 | 4,732 | +2% | 0 | 0 | — |
case-02 | fail→fail | 43,969 | 36,396 | -17% | 1 | 1 | 0% | 6,824 | 6,418 | -6% | 0 | 0 | — |
case-03 | fail→fail | 51,578 | 41,380 | -20% | 1 | 1 | 0% | 6,850 | 6,306 | -8% | 0 | 0 | — |
case-04 | pass→pass | 28,210 | 29,976 | +6% | 1 | 1 | 0% | 4,319 | 5,559 | +29% | 0 | 0 | — |
case-05 | pass→pass | 30,489 | 26,938 | -12% | 1 | 1 | 0% | 4,673 | 5,356 | +15% | 0 | 0 | — |
case-06 | pass→pass | 22,575 | 26,692 | +18% | 1 | 1 | 0% | 3,425 | 4,978 | +45% | 0 | 0 | — |
case-07 | pass→pass | 18,940 | 9,269 | -51% | 1 | 1 | 0% | 2,972 | 2,306 | -22% | 0 | 0 | — |
case-08 | pass→pass | 27,258 | 16,934 | -38% | 1 | 1 | 0% | 3,731 | 3,350 | -10% | 0 | 0 | — |
case-09 | pass→pass | 21,884 | 18,623 | -15% | 1 | 1 | 0% | 3,091 | 3,295 | +7% | 0 | 0 | — |
case-10 | fail→fail | 17,680 | 12,013 | -32% | 1 | 1 | 0% | 2,574 | 2,806 | +9% | 0 | 0 | — |
case-12 | pass→pass | 16,760 | 3,236 | -81% | 1 | 1 | 0% | 2,176 | 1,304 | -40% | 0 | 0 | — |
case-13 | pass→pass | 9,148 | 4,398 | -52% | 1 | 1 | 0% | 1,544 | 1,468 | -5% | 0 | 0 | — |
case-14 | pass→pass | 16,691 | 12,330 | -26% | 1 | 1 | 0% | 2,435 | 2,936 | +21% | 0 | 0 | — |
case-15 | pass→pass | 19,208 | 18,407 | -4% | 1 | 1 | 0% | 2,670 | 3,910 | +46% | 0 | 0 | — |
case-16 | pass→pass | 21,856 | 14,324 | -34% | 1 | 1 | 0% | 2,754 | 2,640 | -4% | 0 | 0 | — |
case-18 | pass→pass | 11,815 | 8,409 | -29% | 1 | 1 | 0% | 1,909 | 1,958 | +3% | 0 | 0 | — |
case-19 | pass→pass | 21,585 | 18,718 | -13% | 1 | 1 | 0% | 2,878 | 3,655 | +27% | 0 | 0 | — |
case-20 | pass→pass | 14,437 | 4,142 | -71% | 1 | 1 | 0% | 1,933 | 1,506 | -22% | 0 | 0 | — |
case-21 | fail→pass | 6,063 | 9,164 | +51% | 1 | 1 | 0% | 803 | 2,046 | +155% | 0 | 0 | — |
case-22 | pass→pass | 18,003 | 15,072 | -16% | 1 | 1 | 0% | 2,644 | 2,572 | -3% | 0 | 0 | — |
case-23 | pass→pass | 24,289 | 12,423 | -49% | 1 | 1 | 0% | 3,372 | 2,606 | -23% | 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.