Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Command execution timing — microsecond precision, statistical analysis, comparison. Zero barrier — use with x-cmd or shell built-in time command. Use for "timing", "benchmark", "performance", "profiling".
.claude/skills/x-cmd-time/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -76% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -59% | 0% |
Measure command execution time. From single-shot microsecond timing to multi-run statistical analysis.
bash# With x-cmd x time <command> # Single microsecond timing x time stress -c 100 <command> # 100-run statistics x time cmp 'cmd1' 'cmd2' 'cmd3' # Compare commands # Without x-cmd — use shell built-ins time <command> # Basic timing # For benchmarking, use a loop: for i in $(seq 1 100); do start=$(date +%s%N) <command> end=$(date +%s%N) echo $(( end - start )) done
x time <cmd> — Single run, microsecond precisionx time getms <cmd> — Millisecond precisionx time stress -c N <cmd> — N runs with full statsx time stress -b N <cmd> — Auto-decide runs within budgetx time cmp 'c1' 'c2' — Compare multiple commands--csv CSV for programmatic analysis--tsv Tab-separated--yml YAMLStarting with the essentials. Will add:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,304 | 4,896 | -47% | 1 | 1 | 0% | 1,382 | 847 | -39% | 0 | 0 | — |
case-02 | fail→pass | 13,200 | 5,814 | -56% | 1 | 1 | 0% | 2,273 | 1,069 | -53% | 0 | 0 | — |
case-03 | fail→pass | 11,907 | 3,733 | -69% | 1 | 1 | 0% | 1,825 | 822 | -55% | 0 | 0 | — |
case-04 | fail→pass | 14,498 | 3,280 | -77% | 1 | 1 | 0% | 2,616 | 632 | -76% | 0 | 0 | — |
case-05 | fail→fail | 8,517 | 4,323 | -49% | 1 | 1 | 0% | 1,430 | 1,017 | -29% | 0 | 0 | — |
case-06 | fail→pass | 9,996 | 3,476 | -65% | 1 | 1 | 0% | 1,973 | 805 | -59% | 0 | 0 | — |
case-07 | pass→pass | 21,374 | 3,308 | -85% | 1 | 1 | 0% | 1,492 | 785 | -47% | 0 | 0 | — |
case-08 | pass→pass | 24,029 | 11,818 | -51% | 1 | 1 | 0% | 4,496 | 2,542 | -43% | 0 | 0 | — |
case-09 | pass→pass | 5,722 | 2,962 | -48% | 1 | 1 | 0% | 686 | 706 | +3% | 0 | 0 | — |
case-10 | fail→pass | 7,915 | 8,009 | +1% | 1 | 1 | 0% | 1,255 | 701 | -44% | 0 | 0 | — |
case-11 | fail→pass | 19,122 | 3,067 | -84% | 1 | 1 | 0% | 1,036 | 689 | -33% | 0 | 0 | — |
case-12 | fail→pass | 11,509 | 3,128 | -73% | 1 | 1 | 0% | 1,938 | 647 | -67% | 0 | 0 | — |
case-13 | pass→pass | 15,766 | 3,039 | -81% | 1 | 1 | 0% | 2,351 | 625 | -73% | 0 | 0 | — |
case-14 | fail→pass | 14,310 | 4,621 | -68% | 1 | 1 | 0% | 2,447 | 984 | -60% | 0 | 0 | — |
case-15 | fail→pass | 7,233 | 3,669 | -49% | 1 | 1 | 0% | 1,169 | 745 | -36% | 0 | 0 | — |
case-16 | fail→pass | 18,748 | 2,948 | -84% | 1 | 1 | 0% | 1,149 | 600 | -48% | 0 | 0 | — |
case-17 | pass→pass | 14,066 | 6,646 | -53% | 1 | 1 | 0% | 1,625 | 1,203 | -26% | 0 | 0 | — |
case-18 | fail→pass | 18,351 | 2,792 | -85% | 1 | 1 | 0% | 1,730 | 674 | -61% | 0 | 0 | — |
case-19 | fail→pass | 27,743 | 3,019 | -89% | 1 | 1 | 0% | 1,972 | 669 | -66% | 0 | 0 | — |
case-20 | pass→pass | 13,838 | 14,095 | +2% | 1 | 1 | 0% | 2,274 | 2,615 | +15% | 0 | 0 | — |
case-21 | pass→pass | 7,667 | 6,671 | -13% | 1 | 1 | 0% | 1,311 | 1,297 | -1% | 0 | 0 | — |
case-22 | pass→pass | 12,332 | 8,116 | -34% | 1 | 1 | 0% | 1,893 | 1,783 | -6% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +59 percentage points is the difference between those two pass rates over the 21 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.