Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.
.claude/skills/hoangnguyen0403-skill-benchmark/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -45% | 0% |
> !IMPORTANT] > Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.
Optional args: slug=<feature>, ticket=<id/url>, mode=interactive|autonomous|channel, channel=<id>, auto_continue=true|false, profile=business|hybrid|technical.
When the user asks to perform this workflow, execute the following steps:
> Goal: Quantify how much active skills improve implementation quality. Deliver a prioritized compliance delta and skill applicability report.
Identify the tech stack and all active skills in AGENTS.md.
bash# 1. Total source files and lines changed find src -name "*.ts" -o -name "*.tsx" | xargs wc -l 2>/dev/null | sort -rn | head -20 # 2. Check active skill registry cat AGENTS.md | head -80
Pick the file automatically. Rank candidates by the severity of anti-patterns:
Source your scorecard from evals/evals.json, not from hardcoded patterns. Follow the Scorecard Rubric in <SKILLS>/common/common-skill-creator/references/benchmark.md when synced:
<SKILLS>/<category>/<skill>/evals/evals.json.pressure_scenarios, rationalizations, red_flags, and behavior_assertions.Output the scorecard and compliant score using the templates in <SKILLS>/common/common-skill-creator/references/benchmark.md when synced.
For every ❌ FAIL, identify the root cause using the Iteration Table in: <SKILLS>/common/common-skill-creator/references/benchmark.md when synced.
Recommend any skills that are noisy or non-applicable for the project.
yamlexclude: - [skill-id] # reason
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,567 | 3,133 | -81% | 1 | 1 | 0% | 3,055 | 1,141 | -63% | 0 | 0 | — |
case-02 | fail→fail | 11,897 | 5,561 | -53% | 1 | 1 | 0% | 2,115 | 1,054 | -50% | 0 | 0 | — |
case-03 | fail→fail | 17,171 | 4,950 | -71% | 1 | 1 | 0% | 2,765 | 1,000 | -64% | 0 | 0 | — |
case-04 | pass→pass | 14,504 | 14,005 | -3% | 1 | 1 | 0% | 2,409 | 3,224 | +34% | 0 | 0 | — |
case-05 | pass→fail | 14,109 | 5,093 | -64% | 1 | 1 | 0% | 3,220 | 993 | -69% | 0 | 0 | — |
case-06 | pass→fail | 8,592 | 4,975 | -42% | 1 | 1 | 0% | 1,859 | 945 | -49% | 0 | 0 | — |
case-07 | fail→pass | 13,475 | 4,148 | -69% | 1 | 1 | 0% | 2,566 | 1,394 | -46% | 0 | 0 | — |
case-08 | fail→pass | 12,565 | 3,451 | -73% | 1 | 1 | 0% | 2,173 | 1,335 | -39% | 0 | 0 | — |
case-09 | fail→pass | 11,508 | 2,328 | -80% | 1 | 1 | 0% | 1,594 | 1,145 | -28% | 0 | 0 | — |
case-10 | fail→pass | 7,392 | 2,796 | -62% | 1 | 1 | 0% | 1,261 | 1,169 | -7% | 0 | 0 | — |
case-11 | fail→pass | 13,447 | 2,703 | -80% | 1 | 1 | 0% | 2,281 | 1,252 | -45% | 0 | 0 | — |
case-12 | fail→pass | 12,810 | 6,539 | -49% | 1 | 1 | 0% | 2,141 | 1,853 | -13% | 0 | 0 | — |
case-13 | fail→fail | 12,648 | 4,740 | -63% | 1 | 1 | 0% | 2,241 | 1,540 | -31% | 0 | 0 | — |
case-14 | fail→pass | 14,672 | 8,117 | -45% | 1 | 1 | 0% | 2,553 | 2,234 | -12% | 0 | 0 | — |
case-15 | fail→pass | 12,665 | 9,576 | -24% | 1 | 1 | 0% | 2,032 | 2,143 | +5% | 0 | 0 | — |
case-16 | fail→pass | 15,741 | 2,366 | -85% | 1 | 1 | 0% | 2,459 | 1,052 | -57% | 0 | 0 | — |
case-17 | pass→pass | 9,860 | 4,374 | -56% | 1 | 1 | 0% | 1,661 | 1,322 | -20% | 0 | 0 | — |
case-18 | pass→pass | 7,931 | 3,526 | -56% | 1 | 1 | 0% | 1,237 | 1,299 | +5% | 0 | 0 | — |
case-19 | pass→pass | 10,258 | 5,485 | -47% | 1 | 1 | 0% | 1,767 | 1,612 | -9% | 0 | 0 | — |
case-20 | fail→pass | 11,255 | 1,663 | -85% | 1 | 1 | 0% | 962 | 963 | +0% | 0 | 0 | — |
case-21 | fail→pass | 8,526 | 2,774 | -67% | 1 | 1 | 0% | 1,501 | 1,133 | -25% | 0 | 0 | — |
case-22 | pass→pass | 11,257 | 5,958 | -47% | 1 | 1 | 0% | 1,831 | 1,705 | -7% | 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 18 counted toward the lift figure. The other 4 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 +41 percentage points is the difference between those two pass rates over the 18 comparable cases. 5 cases got worse with the skill loaded, and they are included in that figure.
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.