Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing stylesheets, component styles, and responsive behavior related to Keep CSS specificity low and flat. Check the rendered layout across breakpoints and interaction states before proposing a fix.
.claude/skills/thedaviddias-specificity-management/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 17% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
| case-16 | ✓→✓ | ＝ Same ✓ | -7% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -30% | 0% |
High specificity creates an escalation problem — once you use an ID selector, you need another ID to override it. Developers respond with !important, which escalates further. Flat, low-specificity CSS is predictable: later rules and more-specific selectors win cleanly, and the cascade works as intended.
Analyze the selectors in this CSS file for specificity issues: ID selectors used for styling, overly nested selectors, and unnecessary !important.
Flatten high-specificity selectors to use classes instead of IDs, reduce nesting depth, and remove unnecessary !important declarations.
Explain CSS specificity scoring, why high specificity causes maintenance problems, and how to keep specificity flat using classes and BEM.
Review stylesheets, component styles, and responsive states related to Keep CSS specificity low and flat. Flag exact selectors, declarations, or breakpoints that violate the rule in the rendered UI.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/css/specificity-management
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 5,001 | 4,446 | -11% | 1 | 1 | 0% | 938 | 1,157 | +23% | 0 | 0 | — |
case-16 | pass→pass | 13,523 | 11,502 | -15% | 1 | 1 | 0% | 2,266 | 2,115 | -7% | 0 | 0 | — |
case-02 | pass→pass | 10,789 | 5,470 | -49% | 1 | 1 | 0% | 1,859 | 1,294 | -30% | 0 | 0 | — |
case-03 | pass→pass | 9,887 | 7,608 | -23% | 1 | 1 | 0% | 1,822 | 1,636 | -10% | 0 | 0 | — |
case-04 | pass→pass | 13,615 | 11,450 | -16% | 1 | 1 | 0% | 2,396 | 2,199 | -8% | 0 | 0 | — |
case-05 | pass→pass | 4,174 | 3,815 | -9% | 1 | 1 | 0% | 685 | 1,063 | +55% | 0 | 0 | — |
case-06 | pass→pass | 11,575 | 7,709 | -33% | 1 | 1 | 0% | 2,054 | 1,722 | -16% | 0 | 0 | — |
case-07 | pass→pass | 11,232 | 10,285 | -8% | 1 | 1 | 0% | 2,342 | 2,321 | -1% | 0 | 0 | — |
case-08 | fail→pass | 10,574 | 6,695 | -37% | 1 | 1 | 0% | 2,026 | 1,630 | -20% | 0 | 0 | — |
case-09 | pass→pass | 10,128 | 6,481 | -36% | 1 | 1 | 0% | 1,948 | 1,637 | -16% | 0 | 0 | — |
case-10 | pass→pass | 9,529 | 7,283 | -24% | 1 | 1 | 0% | 1,990 | 1,691 | -15% | 0 | 0 | — |
case-11 | pass→pass | 5,406 | 4,683 | -13% | 1 | 1 | 0% | 1,138 | 1,239 | +9% | 0 | 0 | — |
case-12 | pass→pass | 9,892 | 8,726 | -12% | 1 | 1 | 0% | 1,858 | 1,976 | +6% | 0 | 0 | — |
case-13 | pass→pass | 7,461 | 4,029 | -46% | 1 | 1 | 0% | 1,349 | 1,015 | -25% | 0 | 0 | — |
case-14 | pass→pass | 8,766 | 5,689 | -35% | 1 | 1 | 0% | 1,616 | 1,387 | -14% | 0 | 0 | — |
case-15 | pass→pass | 9,213 | 7,196 | -22% | 1 | 1 | 0% | 1,690 | 1,474 | -13% | 0 | 0 | — |
case-17 | pass→pass | 10,659 | 6,395 | -40% | 1 | 1 | 0% | 1,966 | 1,473 | -25% | 0 | 0 | — |
case-18 | pass→pass | 14,913 | 13,947 | -6% | 1 | 1 | 0% | 2,538 | 2,814 | +11% | 0 | 0 | — |
case-19 | pass→pass | 10,897 | 7,668 | -30% | 1 | 1 | 0% | 2,083 | 1,750 | -16% | 0 | 0 | — |
case-20 | pass→fail | 11,260 | 11,992 | +7% | 1 | 1 | 0% | 2,056 | 2,400 | +17% | 0 | 0 | — |
case-21 | pass→pass | 8,462 | 7,816 | -8% | 1 | 1 | 0% | 1,526 | 1,660 | +9% | 0 | 0 | — |
case-22 | pass→pass | 6,234 | 5,644 | -9% | 1 | 1 | 0% | 1,082 | 1,343 | +24% | 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. The headline lift of 0 percentage points is the difference between those two pass rates over the 22 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.