Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →k6 load testing expertise for performance validation and analysis
.claude/skills/a5c-ai-k6-performance-testing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 72% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -70% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -30% | 0% |
This skill provides expert-level capabilities for k6-based performance testing, enabling load test scripting, execution, metrics analysis, and integration with monitoring systems.
performance-testing.js - Performance test implementationapi-testing.js - API performance validationcontinuous-testing.js - CI/CD performance gatesk6 - Load testing tooljavascript{ kind: 'skill', skill: { name: 'k6-performance', context: { action: 'execute-load-test', script: 'tests/performance/load.js', vus: 100, duration: '5m', thresholds: { 'http_req_duration': ['p(95)<500'] } } } }
The skill can execute k6 scripts locally or configure cloud execution for distributed load testing.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 5,903 | 14,573 | +147% | 1 | 1 | 0% | 1,140 | 1,958 | +72% | 0 | 0 | — |
case-02 | fail→pass | 12,038 | 1,715 | -86% | 1 | 1 | 0% | 2,195 | 669 | -70% | 0 | 0 | — |
case-03 | fail→pass | 8,187 | 2,409 | -71% | 1 | 1 | 0% | 1,350 | 833 | -38% | 0 | 0 | — |
case-04 | fail→fail | 16,886 | 11,005 | -35% | 1 | 1 | 0% | 3,126 | 2,676 | -14% | 0 | 0 | — |
case-05 | fail→pass | 7,936 | 1,975 | -75% | 1 | 1 | 0% | 1,295 | 759 | -41% | 0 | 0 | — |
case-06 | pass→pass | 7,704 | 5,071 | -34% | 1 | 1 | 0% | 1,146 | 1,250 | +9% | 0 | 0 | — |
case-07 | pass→pass | 6,622 | 4,342 | -34% | 1 | 1 | 0% | 1,256 | 1,220 | -3% | 0 | 0 | — |
case-08 | pass→pass | 4,517 | 2,854 | -37% | 1 | 1 | 0% | 958 | 1,037 | +8% | 0 | 0 | — |
case-09 | pass→pass | 7,435 | 3,481 | -53% | 1 | 1 | 0% | 1,173 | 1,034 | -12% | 0 | 0 | — |
case-10 | pass→pass | 5,912 | 2,699 | -54% | 1 | 1 | 0% | 1,023 | 686 | -33% | 0 | 0 | — |
case-11 | pass→pass | 20,326 | 12,972 | -36% | 1 | 1 | 0% | 2,029 | 2,595 | +28% | 0 | 0 | — |
case-12 | pass→pass | 4,187 | 4,555 | +9% | 1 | 1 | 0% | 701 | 1,282 | +83% | 0 | 0 | — |
case-13 | pass→pass | 7,662 | 10,221 | +33% | 1 | 1 | 0% | 1,380 | 2,346 | +70% | 0 | 0 | — |
case-14 | pass→pass | 5,877 | 4,040 | -31% | 1 | 1 | 0% | 1,138 | 1,218 | +7% | 0 | 0 | — |
case-15 | pass→pass | 12,749 | 6,089 | -52% | 1 | 1 | 0% | 1,974 | 1,575 | -20% | 0 | 0 | — |
case-16 | fail→pass | 12,675 | 5,998 | -53% | 1 | 1 | 0% | 2,038 | 1,436 | -30% | 0 | 0 | — |
case-17 | pass→pass | 10,663 | 8,946 | -16% | 1 | 1 | 0% | 1,658 | 2,026 | +22% | 0 | 0 | — |
case-18 | pass→pass | 17,396 | 15,502 | -11% | 1 | 1 | 0% | 2,479 | 2,676 | +8% | 0 | 0 | — |
case-19 | pass→pass | 11,510 | 11,217 | -3% | 1 | 1 | 0% | 1,896 | 2,621 | +38% | 0 | 0 | — |
case-20 | fail→fail | 12,384 | 9,835 | -21% | 1 | 1 | 0% | 2,675 | 2,321 | -13% | 0 | 0 | — |
case-21 | fail→fail | 7,473 | 4,578 | -39% | 1 | 1 | 0% | 1,116 | 1,354 | +21% | 0 | 0 | — |
case-22 | fail→fail | 17,735 | 27,152 | +53% | 1 | 1 | 0% | 3,378 | 3,345 | -1% | 0 | 0 | — |
case-23 | fail→fail | 10,410 | 9,703 | -7% | 1 | 1 | 0% | 1,861 | 2,433 | +31% | 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 +22 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.