Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use to evaluate or design a user interaction flow — states, edge cases, microinteractions, motion, and feedback patterns.
.claude/skills/getcrew44-interaction-review/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 39% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 96% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 42% | 0% |
Make sure every state a user can land in has been considered, and that motion serves the user rather than decorating.
For each user-facing surface, walk through:
| State | Question | |-------|----------| | Loading | What does the user see while waiting? Is it clear progress is happening? | | Empty | What does the user see with zero results, zero items, zero history? Does the empty state guide the next action? | | Error | When the network or backend fails, can the user understand what happened and recover? | | Success | Is feedback clear and the next action obvious? | | Partial | When the operation half-succeeded, is the partial result accurately reported? |
Motion should communicate, not decorate. Every animation answers one of: confirm an action occurred, orient where things came from or go to, focus attention on a change, or preserve context during a transition. If it does none of these, cut it.
Use a timing scale:
Prefer spring or ease-out for entrances, ease-in for exits, transform/opacity for performance. Honor prefers-reduced-motion.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,177 | 25,002 | +24% | 1 | 1 | 0% | 2,862 | 4,133 | +44% | 0 | 0 | — |
case-02 | fail→fail | 17,952 | 23,464 | +31% | 1 | 1 | 0% | 2,669 | 3,844 | +44% | 0 | 0 | — |
case-03 | pass→pass | 22,975 | 31,454 | +37% | 1 | 1 | 0% | 4,795 | 6,681 | +39% | 0 | 0 | — |
case-04 | pass→pass | 12,116 | 20,744 | +71% | 1 | 1 | 0% | 1,905 | 3,734 | +96% | 0 | 0 | — |
case-05 | pass→pass | 17,794 | 23,648 | +33% | 1 | 1 | 0% | 3,420 | 4,871 | +42% | 0 | 0 | — |
case-06 | pass→pass | 12,639 | 14,423 | +14% | 1 | 1 | 0% | 1,758 | 2,696 | +53% | 0 | 0 | — |
case-07 | pass→pass | 9,813 | 10,939 | +11% | 1 | 1 | 0% | 1,394 | 2,177 | +56% | 0 | 0 | — |
case-08 | pass→pass | 13,890 | 14,623 | +5% | 1 | 1 | 0% | 2,099 | 2,665 | +27% | 0 | 0 | — |
case-09 | pass→pass | 11,741 | 14,570 | +24% | 1 | 1 | 0% | 1,722 | 2,609 | +52% | 0 | 0 | — |
case-10 | fail→fail | 16,449 | 18,715 | +14% | 1 | 1 | 0% | 2,579 | 3,096 | +20% | 0 | 0 | — |
case-11 | pass→pass | 14,155 | 17,651 | +25% | 1 | 1 | 0% | 2,261 | 3,048 | +35% | 0 | 0 | — |
case-12 | fail→fail | 11,610 | 16,954 | +46% | 1 | 1 | 0% | 1,794 | 2,945 | +64% | 0 | 0 | — |
case-13 | pass→pass | 10,197 | 8,015 | -21% | 1 | 1 | 0% | 1,687 | 1,742 | +3% | 0 | 0 | — |
case-14 | pass→pass | 8,099 | 6,279 | -22% | 1 | 1 | 0% | 1,278 | 1,564 | +22% | 0 | 0 | — |
case-15 | fail→pass | 11,192 | 8,122 | -27% | 1 | 1 | 0% | 1,911 | 1,740 | -9% | 0 | 0 | — |
case-16 | pass→pass | 10,379 | 10,056 | -3% | 1 | 1 | 0% | 1,764 | 2,117 | +20% | 0 | 0 | — |
case-17 | pass→pass | 12,297 | 13,212 | +7% | 1 | 1 | 0% | 2,076 | 2,830 | +36% | 0 | 0 | — |
case-18 | pass→pass | 11,350 | 11,272 | -1% | 1 | 1 | 0% | 1,742 | 2,333 | +34% | 0 | 0 | — |
case-19 | fail→fail | 14,713 | 13,876 | -6% | 1 | 1 | 0% | 2,185 | 2,609 | +19% | 0 | 0 | — |
case-20 | pass→pass | 12,230 | 15,050 | +23% | 1 | 1 | 0% | 1,977 | 2,657 | +34% | 0 | 0 | — |
case-21 | pass→pass | 10,537 | 12,659 | +20% | 1 | 1 | 0% | 1,542 | 2,467 | +60% | 0 | 0 | — |
case-22 | pass→pass | 15,310 | 21,327 | +39% | 1 | 1 | 0% | 2,122 | 3,441 | +62% | 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 +9 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.