Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Dead code cleanup and consolidation specialist. Use PROACTIVELY for removing unused code, duplicates, and refactoring. Runs analysis tools (knip, depcheck, ts-prune) to identify dead code and safely removes it.
.claude/skills/kunanonj-agent-refactor-cleaner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -62% | 0% |
You are an expert refactoring specialist focused on code cleanup and consolidation. Your mission is to identify and remove dead code, duplicates, and unused exports.
bashnpx knip # Unused files, exports, dependencies npx depcheck # Unused npm dependencies npx ts-prune # Unused TypeScript exports npx eslint . --report-unused-disable-directives # Unused eslint directives
For each item to remove:
Before removing:
After each batch:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 12,802 | 10,092 | -21% | 1 | 1 | 0% | 2,447 | 2,411 | -1% | 0 | 0 | — |
case-02 | fail→fail | 1,673 | 8,176 | +389% | 1 | 1 | 0% | 232 | 2,042 | +780% | 0 | 0 | — |
case-03 | fail→pass | 14,664 | 7,981 | -46% | 1 | 1 | 0% | 2,512 | 1,962 | -22% | 0 | 0 | — |
case-04 | pass→pass | 6,348 | 3,582 | -44% | 1 | 1 | 0% | 1,063 | 1,290 | +21% | 0 | 0 | — |
case-05 | pass→pass | 9,390 | 4,606 | -51% | 1 | 1 | 0% | 1,515 | 1,514 | -0% | 0 | 0 | — |
case-06 | pass→pass | 7,515 | 5,490 | -27% | 1 | 1 | 0% | 1,264 | 1,642 | +30% | 0 | 0 | — |
case-07 | pass→pass | 7,886 | 5,332 | -32% | 1 | 1 | 0% | 1,303 | 1,651 | +27% | 0 | 0 | — |
case-08 | fail→pass | 12,345 | 2,222 | -82% | 1 | 1 | 0% | 1,947 | 1,110 | -43% | 0 | 0 | — |
case-09 | fail→pass | 9,538 | 3,709 | -61% | 1 | 1 | 0% | 1,480 | 1,320 | -11% | 0 | 0 | — |
case-10 | pass→pass | 15,231 | 12,558 | -18% | 1 | 1 | 0% | 2,250 | 2,641 | +17% | 0 | 0 | — |
case-11 | pass→pass | 10,826 | 6,994 | -35% | 1 | 1 | 0% | 1,611 | 1,851 | +15% | 0 | 0 | — |
case-12 | fail→fail | 11,876 | 8,046 | -32% | 1 | 1 | 0% | 1,901 | 2,179 | +15% | 0 | 0 | — |
case-13 | pass→pass | 12,585 | 7,013 | -44% | 1 | 1 | 0% | 1,822 | 1,868 | +3% | 0 | 0 | — |
case-14 | pass→pass | 13,718 | 9,447 | -31% | 1 | 1 | 0% | 2,432 | 2,225 | -9% | 0 | 0 | — |
case-15 | pass→pass | 11,318 | 5,005 | -56% | 1 | 1 | 0% | 1,998 | 1,551 | -22% | 0 | 0 | — |
case-16 | pass→pass | 11,542 | 2,655 | -77% | 1 | 1 | 0% | 1,995 | 1,254 | -37% | 0 | 0 | — |
case-17 | pass→pass | 7,983 | 3,695 | -54% | 1 | 1 | 0% | 1,165 | 1,438 | +23% | 0 | 0 | — |
case-18 | pass→pass | 10,994 | 3,690 | -66% | 1 | 1 | 0% | 1,671 | 1,408 | -16% | 0 | 0 | — |
case-19 | pass→pass | 10,984 | 10,697 | -3% | 1 | 1 | 0% | 1,901 | 2,347 | +23% | 0 | 0 | — |
case-20 | pass→fail | 17,505 | 6,845 | -61% | 1 | 1 | 0% | 4,661 | 1,780 | -62% | 0 | 0 | — |
case-21 | pass→pass | 18,183 | 12,071 | -34% | 1 | 1 | 0% | 3,669 | 3,293 | -10% | 0 | 0 | — |
case-22 | pass→pass | 14,646 | 12,847 | -12% | 1 | 1 | 0% | 2,483 | 3,004 | +21% | 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 +14 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.