Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Interactive technical design quality review and validation
.claude/skills/bilal140202-kiro-validate-design/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 48% | 0% |
<background_information>
</background_information>
<instructions>
Interactive design quality review for feature $1 based on approved requirements and design document.
{{KIRO_DIR}}/specs/$1/spec.json for language and metadata{{KIRO_DIR}}/specs/$1/requirements.md for requirements{{KIRO_DIR}}/specs/$1/design.md for design documentproduct.md, tech.md, structure.mdThe following research areas are independent and can be executed in parallel:
rules/design-review.md from this skill's directory for review criteriaIf multi-agent is enabled, spawn sub-agents for each area above. Otherwise execute sequentially.
After all parallel research completes, synthesize findings for review.
</instructions>
Provide output in the language specified in spec.json with:
Format Requirements:
/kiro-spec-design $1 first to generate design document"en) if spec.json doesn't specify languageIf Design Passes Validation (GO Decision):
/kiro-spec-tasks $1 to generate implementation tasks/kiro-spec-tasks $1 -y to auto-approve and proceed directlyIf Design Needs Revision (NO-GO Decision):
/kiro-spec-design $1 with improvements/kiro-validate-design $1Note: Design validation is recommended but optional. Quality review helps catch issues early.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 9,095 | 12,221 | +34% | 1 | 1 | 0% | 1,479 | 1,327 | -10% | 0 | 0 | — |
case-02 | fail→fail | 5,188 | 4,299 | -17% | 1 | 1 | 0% | 947 | 1,383 | +46% | 0 | 0 | — |
case-03 | fail→fail | 13,068 | 5,008 | -62% | 1 | 1 | 0% | 2,188 | 1,337 | -39% | 0 | 0 | — |
case-04 | fail→fail | 25,897 | 5,772 | -78% | 1 | 1 | 0% | 6,175 | 1,378 | -78% | 0 | 0 | — |
case-05 | fail→fail | 16,738 | 6,560 | -61% | 1 | 1 | 0% | 2,848 | 1,421 | -50% | 0 | 0 | — |
case-06 | fail→fail | 14,921 | 9,462 | -37% | 1 | 1 | 0% | 3,105 | 1,819 | -41% | 0 | 0 | — |
case-07 | fail→pass | 4,219 | 1,830 | -57% | 1 | 1 | 0% | 829 | 1,403 | +69% | 0 | 0 | — |
case-08 | fail→fail | 19,686 | 23,262 | +18% | 1 | 1 | 0% | 3,442 | 1,344 | -61% | 0 | 0 | — |
case-09 | fail→fail | 9,228 | 4,312 | -53% | 1 | 1 | 0% | 1,675 | 1,459 | -13% | 0 | 0 | — |
case-10 | fail→fail | 13,878 | 4,854 | -65% | 1 | 1 | 0% | 2,697 | 1,501 | -44% | 0 | 0 | — |
case-11 | pass→fail | 14,085 | 3,897 | -72% | 1 | 1 | 0% | 2,529 | 1,346 | -47% | 0 | 0 | — |
case-12 | fail→fail | 7,502 | 18,787 | +150% | 1 | 1 | 0% | 1,386 | 3,861 | +179% | 0 | 0 | — |
case-13 | pass→pass | 4,304 | 1,921 | -55% | 1 | 1 | 0% | 730 | 1,434 | +96% | 0 | 0 | — |
case-14 | fail→pass | 9,974 | 1,341 | -87% | 1 | 1 | 0% | 1,864 | 1,356 | -27% | 0 | 0 | — |
case-15 | fail→pass | 8,576 | 2,192 | -74% | 1 | 1 | 0% | 1,699 | 1,487 | -12% | 0 | 0 | — |
case-16 | fail→pass | 18,918 | 1,987 | -89% | 1 | 1 | 0% | 1,044 | 1,444 | +38% | 0 | 0 | — |
case-17 | fail→pass | 7,326 | 2,594 | -65% | 1 | 1 | 0% | 1,017 | 1,505 | +48% | 0 | 0 | — |
case-18 | fail→pass | 8,000 | 1,575 | -80% | 1 | 1 | 0% | 1,283 | 1,336 | +4% | 0 | 0 | — |
case-19 | pass→pass | 15,544 | 4,547 | -71% | 1 | 1 | 0% | 2,635 | 1,978 | -25% | 0 | 0 | — |
case-20 | pass→pass | 4,552 | 1,744 | -62% | 1 | 1 | 0% | 750 | 1,398 | +86% | 0 | 0 | — |
case-21 | fail→pass | 5,950 | 1,736 | -71% | 1 | 1 | 0% | 1,097 | 1,336 | +22% | 0 | 0 | — |
case-22 | fail→pass | 11,677 | 1,970 | -83% | 1 | 1 | 0% | 2,373 | 1,446 | -39% | 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 10 counted toward the lift figure. The other 12 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 +32 percentage points is the difference between those two pass rates over the 10 comparable cases. 1 case got worse with the skill loaded, and it is 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.