Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Refactoring pass — extracts repeated Tailwind class combos and markup into components, and lifts magic values into design tokens. Use when the codebase has obvious duplication, hardcoded hex values or pixel sizes, or when the user says "clean this up" / "extract components" / "tokenize styles". Invoke when the user asks for extract on their UI, or mentions 'extract' alongside design / UI / frontend work.
.claude/skills/educlopez-extract/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 74% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 46% | 0% |
<!-- HARNESS MIRROR — do not edit here. Canonical source: skills/ or commands/. After editing source, copy into cli/assets/<harness>/ and repo-root harness mirrors. -->
Context: this sub-skill is one lens of the broader ui-craft skill. If the ui-craft skill is also installed, read its SKILL.md first for Discovery + Anti-Slop + Craft Test, then apply the specific lens below.
Extract reusable pieces from $ARGUMENTS. Load the ui-craft skill.
Step 1 — Scan the target for extraction candidates:
cva/tv variant if the project uses one — check first).theme.extend.Step 2 — Match the project's conventions, don't impose new ones:
components/ui/, ui/, primitives/, src/components/ before creating a new directory. Reuse.--color-*, --space-*, --font-*, --radius-*). Don't introduce a parallel system.Step 3 — Anti-Slop Test on every new component (from SKILL.md):
MetricCard / PricingTier / InviteRow beats Card / Box / Item.Step 4 — Update every call-site. No half-migrations.
Knob-agnostic — DRY is not tunable.
References to read: references/layout.md (composition patterns), references/typography.md and references/color.md (token naming), plus the project's existing component library (scan before writing).
Output:
| File | Before | After | | --- | --- | --- |
Next step: /tokens — lift the values you just centralized into the token spine (rung 2).
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 28,176 | 4,864 | -83% | 1 | 1 | 0% | 6,225 | 921 | -85% | 0 | 0 | — |
case-02 | fail→fail | 21,119 | 21,198 | +0% | 1 | 1 | 0% | 4,780 | 5,165 | +8% | 0 | 0 | — |
case-03 | fail→pass | 13,721 | 17,331 | +26% | 1 | 1 | 0% | 3,014 | 3,659 | +21% | 0 | 0 | — |
case-04 | pass→pass | 6,201 | 8,895 | +43% | 1 | 1 | 0% | 1,304 | 2,318 | +78% | 0 | 0 | — |
case-05 | pass→pass | 8,884 | 4,320 | -51% | 1 | 1 | 0% | 1,498 | 1,448 | -3% | 0 | 0 | — |
case-06 | pass→pass | 7,912 | 6,356 | -20% | 1 | 1 | 0% | 1,610 | 1,950 | +21% | 0 | 0 | — |
case-07 | fail→pass | 5,774 | 7,666 | +33% | 1 | 1 | 0% | 1,129 | 1,959 | +74% | 0 | 0 | — |
case-08 | pass→pass | 9,383 | 3,633 | -61% | 1 | 1 | 0% | 1,656 | 1,243 | -25% | 0 | 0 | — |
case-09 | pass→pass | 7,741 | 3,715 | -52% | 1 | 1 | 0% | 1,313 | 1,249 | -5% | 0 | 0 | — |
case-10 | pass→pass | 7,783 | 2,372 | -70% | 1 | 1 | 0% | 1,289 | 1,027 | -20% | 0 | 0 | — |
case-11 | pass→pass | 2,075 | 2,389 | +15% | 1 | 1 | 0% | 345 | 1,006 | +192% | 0 | 0 | — |
case-12 | fail→fail | 8,597 | 3,198 | -63% | 1 | 1 | 0% | 1,367 | 1,149 | -16% | 0 | 0 | — |
case-13 | fail→fail | 6,515 | 1,638 | -75% | 1 | 1 | 0% | 1,018 | 879 | -14% | 0 | 0 | — |
case-14 | fail→pass | 8,343 | 8,183 | -2% | 1 | 1 | 0% | 1,623 | 2,387 | +47% | 0 | 0 | — |
case-15 | pass→pass | 8,979 | 5,918 | -34% | 1 | 1 | 0% | 1,767 | 1,696 | -4% | 0 | 0 | — |
case-16 | pass→pass | 10,182 | 6,245 | -39% | 1 | 1 | 0% | 2,082 | 1,794 | -14% | 0 | 0 | — |
case-17 | fail→pass | 7,151 | 6,662 | -7% | 1 | 1 | 0% | 1,310 | 1,867 | +43% | 0 | 0 | — |
case-18 | fail→pass | 4,602 | 3,066 | -33% | 1 | 1 | 0% | 839 | 1,225 | +46% | 0 | 0 | — |
case-19 | fail→pass | 2,875 | 1,964 | -32% | 1 | 1 | 0% | 487 | 930 | +91% | 0 | 0 | — |
case-20 | pass→pass | 10,345 | 16,237 | +57% | 1 | 1 | 0% | 2,165 | 3,597 | +66% | 0 | 0 | — |
case-21 | pass→pass | 8,246 | 7,301 | -11% | 1 | 1 | 0% | 1,748 | 2,166 | +24% | 0 | 0 | — |
case-22 | pass→pass | 16,756 | 23,073 | +38% | 1 | 1 | 0% | 4,077 | 5,361 | +31% | 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 +27 percentage points is the difference between those two pass rates over the 21 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.