Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Comprehensive network testing, benchmarking, and performance validation skill
.claude/skills/a5c-ai-network-testing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -51% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 74% | 0% |
Comprehensive skill for network testing, benchmarking, and performance validation across all network layers.
iperf3 - Network bandwidth measurementnetperf - Network performance testingwrk - HTTP benchmarking toolk6 - Modern load testing toolhey - HTTP load generatortc - Traffic control for network shapingmtr - Network diagnostic toolhping3 - TCP/IP packet assemblerbashiperf3 -c server.example.com -t 30 -P 4 iperf3 -s -p 5201
bashwrk -t12 -c400 -d30s http://localhost:8080/ k6 run --vus 100 --duration 30s script.js hey -n 10000 -c 100 http://localhost:8080/
bashmtr --report --report-cycles 100 example.com hping3 -S -p 80 -c 100 example.com
bashtc qdisc add dev eth0 root netem delay 100ms 10ms tc qdisc add dev eth0 root netem loss 5%
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 2,786 | 14,462 | +419% | 1 | 1 | 0% | 518 | 899 | +74% | 0 | 0 | — |
case-02 | pass→pass | 3,290 | 3,167 | -4% | 1 | 1 | 0% | 492 | 1,145 | +133% | 0 | 0 | — |
case-03 | pass→pass | 4,886 | 3,195 | -35% | 1 | 1 | 0% | 998 | 1,150 | +15% | 0 | 0 | — |
case-04 | pass→pass | 5,120 | 1,906 | -63% | 1 | 1 | 0% | 784 | 934 | +19% | 0 | 0 | — |
case-05 | pass→pass | 4,339 | 2,485 | -43% | 1 | 1 | 0% | 793 | 1,001 | +26% | 0 | 0 | — |
case-06 | pass→pass | 5,398 | 3,222 | -40% | 1 | 1 | 0% | 980 | 1,104 | +13% | 0 | 0 | — |
case-07 | pass→pass | 8,149 | 3,105 | -62% | 1 | 1 | 0% | 1,329 | 1,148 | -14% | 0 | 0 | — |
case-08 | pass→pass | 6,496 | 4,144 | -36% | 1 | 1 | 0% | 1,337 | 1,266 | -5% | 0 | 0 | — |
case-09 | pass→pass | 3,652 | 2,616 | -28% | 1 | 1 | 0% | 673 | 993 | +48% | 0 | 0 | — |
case-10 | fail→pass | 34,945 | 18,016 | -48% | 1 | 1 | 0% | 6,998 | 4,385 | -37% | 0 | 0 | — |
case-11 | fail→pass | 19,306 | 1,835 | -90% | 1 | 1 | 0% | 1,346 | 804 | -40% | 0 | 0 | — |
case-12 | fail→pass | 11,722 | 1,895 | -84% | 1 | 1 | 0% | 1,692 | 829 | -51% | 0 | 0 | — |
case-13 | pass→pass | 8,459 | 7,725 | -9% | 1 | 1 | 0% | 1,292 | 1,806 | +40% | 0 | 0 | — |
case-14 | pass→pass | 11,044 | 2,699 | -76% | 1 | 1 | 0% | 1,726 | 892 | -48% | 0 | 0 | — |
case-15 | pass→pass | 14,498 | 8,355 | -42% | 1 | 1 | 0% | 2,199 | 1,928 | -12% | 0 | 0 | — |
case-16 | pass→pass | 19,210 | 17,858 | -7% | 1 | 1 | 0% | 2,942 | 3,303 | +12% | 0 | 0 | — |
case-17 | pass→pass | 14,476 | 6,131 | -58% | 1 | 1 | 0% | 1,986 | 1,474 | -26% | 0 | 0 | — |
case-18 | pass→pass | 16,425 | 17,015 | +4% | 1 | 1 | 0% | 2,842 | 3,381 | +19% | 0 | 0 | — |
case-19 | pass→pass | 8,066 | 5,023 | -38% | 1 | 1 | 0% | 1,518 | 1,491 | -2% | 0 | 0 | — |
case-20 | fail→pass | 7,756 | 3,934 | -49% | 1 | 1 | 0% | 1,340 | 1,229 | -8% | 0 | 0 | — |
case-21 | pass→pass | 7,090 | 5,550 | -22% | 1 | 1 | 0% | 1,466 | 1,601 | +9% | 0 | 0 | — |
case-22 | pass→pass | 5,640 | 6,947 | +23% | 1 | 1 | 0% | 1,120 | 1,998 | +78% | 0 | 0 | — |
case-23 | pass→pass | 5,344 | 5,319 | -0% | 1 | 1 | 0% | 1,124 | 1,693 | +51% | 0 | 0 | — |
case-24 | pass→pass | 4,352 | 2,809 | -35% | 1 | 1 | 0% | 822 | 1,032 | +26% | 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. 24 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 24 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.