Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Decides when quality matters vs move fast, based on Dylan Field (Figma) craft philosophy and Brian Chesky (Airbnb) details obsession. Use when balancing shipping speed with excellence, deciding if refactoring is needed, or determining which details create moats vs which to skip.
.claude/skills/bilal140202-quality-speed/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -16% | 0% |
Claude uses this skill when:
When Quality Matters:
When Speed Matters:
│ USER-FACING │ INTERNAL
────────────────────┼─────────────┼──────────
CORE PRODUCT │ HIGH CRAFT │ MEDIUM
NON-CORE FEATURE │ MEDIUM │ LOW
EXPERIMENT │ LOW │ LOWmarkdown# Feature: [Name] ## Context - User-facing: [yes/no] - Core product loop: [yes/no] - Frequency of use: [daily/weekly/monthly] - Competitive advantage: [yes/no] ## Quality Level Decision **HIGH CRAFT:** - Time investment: [X days] - Polish areas: [list] **MOVE FAST:** - Ship threshold: [works, looks okay] - Time budget: [X days] ## Decision: [HIGH/MEDIUM/LOW craft]
High Craft Signals:
Move Fast Signals:
Dylan Field: > "AI makes design, craft, and quality the new moat for startups."
Brian Chesky: > "Leaders are in the details, but only the details that matter."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,275 | 6,949 | -51% | 1 | 1 | 0% | 2,358 | 1,565 | -34% | 0 | 0 | — |
case-02 | pass→pass | 13,368 | 11,312 | -15% | 1 | 1 | 0% | 2,233 | 2,295 | +3% | 0 | 0 | — |
case-03 | pass→pass | 13,712 | 7,538 | -45% | 1 | 1 | 0% | 2,057 | 1,839 | -11% | 0 | 0 | — |
case-04 | fail→pass | 20,167 | 8,560 | -58% | 1 | 1 | 0% | 3,009 | 1,764 | -41% | 0 | 0 | — |
case-05 | fail→pass | 13,082 | 7,675 | -41% | 1 | 1 | 0% | 2,182 | 1,755 | -20% | 0 | 0 | — |
case-06 | fail→fail | 12,734 | 5,457 | -57% | 1 | 1 | 0% | 2,140 | 1,429 | -33% | 0 | 0 | — |
case-07 | pass→pass | 15,122 | 13,183 | -13% | 1 | 1 | 0% | 2,594 | 2,600 | +0% | 0 | 0 | — |
case-08 | pass→pass | 15,103 | 11,569 | -23% | 1 | 1 | 0% | 2,316 | 2,381 | +3% | 0 | 0 | — |
case-09 | fail→pass | 11,880 | 6,017 | -49% | 1 | 1 | 0% | 1,988 | 1,601 | -19% | 0 | 0 | — |
case-10 | pass→pass | 8,633 | 6,281 | -27% | 1 | 1 | 0% | 1,428 | 1,497 | +5% | 0 | 0 | — |
case-11 | fail→pass | 14,460 | 8,605 | -40% | 1 | 1 | 0% | 2,315 | 1,942 | -16% | 0 | 0 | — |
case-12 | pass→pass | 12,751 | 8,242 | -35% | 1 | 1 | 0% | 2,102 | 1,867 | -11% | 0 | 0 | — |
case-13 | fail→pass | 17,785 | 7,518 | -58% | 1 | 1 | 0% | 3,022 | 1,741 | -42% | 0 | 0 | — |
case-14 | fail→pass | 13,949 | 10,094 | -28% | 1 | 1 | 0% | 2,217 | 2,029 | -8% | 0 | 0 | — |
case-15 | fail→pass | 13,544 | 7,941 | -41% | 1 | 1 | 0% | 2,207 | 1,802 | -18% | 0 | 0 | — |
case-16 | pass→pass | 8,311 | 5,771 | -31% | 1 | 1 | 0% | 1,402 | 1,355 | -3% | 0 | 0 | — |
case-17 | pass→pass | 14,886 | 7,354 | -51% | 1 | 1 | 0% | 2,585 | 1,623 | -37% | 0 | 0 | — |
case-18 | fail→fail | 13,564 | 7,743 | -43% | 1 | 1 | 0% | 2,291 | 1,674 | -27% | 0 | 0 | — |
case-19 | pass→pass | 7,731 | 5,111 | -34% | 1 | 1 | 0% | 1,578 | 1,532 | -3% | 0 | 0 | — |
case-20 | pass→pass | 13,230 | 12,586 | -5% | 1 | 1 | 0% | 2,450 | 2,712 | +11% | 0 | 0 | — |
case-21 | pass→pass | 14,173 | 9,261 | -35% | 1 | 1 | 0% | 2,471 | 2,320 | -6% | 0 | 0 | — |
case-22 | fail→fail | 13,556 | 4,975 | -63% | 1 | 1 | 0% | 2,498 | 1,325 | -47% | 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 +36 percentage points is the difference between those two pass rates over the 22 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.