Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Inspect and reclaim the worktrees Smithers created for <Worktree> lanes. Run `smithers worktree --help` for usage details.
.claude/skills/smithersai-smithers-worktree/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | -91% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -71% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -80% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -81% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 33% | 0% |
List every worktree Smithers created in this repository and the run that owns it.
Remove the worktrees of runs that are over (finished, failed, or cancelled).
| Flag | Type | Default | Description | |------|------|---------|-------------| | --run | string | | Only prune worktrees owned by this run id | | --olderThan | string | | Only prune worktrees untouched for at least this long, e.g. 24h | | --dryRun | boolean | false | Report what would be removed without removing anything | | --force | boolean | false | Also remove worktrees holding uncommitted or unpushed work |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→pass | 28,782 | 1,932 | -93% | 1 | 1 | 0% | 4,946 | 456 | -91% | 0 | 0 | — |
case-07 | pass→pass | 4,020 | 1,850 | -54% | 1 | 1 | 0% | 647 | 418 | -35% | 0 | 0 | — |
case-08 | fail→pass | 10,580 | 2,303 | -78% | 1 | 1 | 0% | 1,835 | 526 | -71% | 0 | 0 | — |
case-01 | fail→pass | 13,344 | 1,819 | -86% | 1 | 1 | 0% | 2,170 | 435 | -80% | 0 | 0 | — |
case-02 | fail→pass | 15,586 | 2,797 | -82% | 1 | 1 | 0% | 3,061 | 592 | -81% | 0 | 0 | — |
case-03 | fail→pass | 5,638 | 1,801 | -68% | 1 | 1 | 0% | 349 | 465 | +33% | 0 | 0 | — |
case-04 | pass→pass | 2,228 | 2,109 | -5% | 1 | 1 | 0% | 418 | 526 | +26% | 0 | 0 | — |
case-05 | pass→fail | 6,218 | 4,736 | -24% | 1 | 1 | 0% | 1,160 | 991 | -15% | 0 | 0 | — |
case-06 | pass→pass | 7,972 | 5,516 | -31% | 1 | 1 | 0% | 1,383 | 1,174 | -15% | 0 | 0 | — |
case-09 | fail→pass | 7,735 | 2,104 | -73% | 1 | 1 | 0% | 1,347 | 526 | -61% | 0 | 0 | — |
case-10 | fail→pass | 5,311 | 1,999 | -62% | 1 | 1 | 0% | 897 | 531 | -41% | 0 | 0 | — |
case-11 | pass→pass | 9,797 | 1,852 | -81% | 1 | 1 | 0% | 1,592 | 478 | -70% | 0 | 0 | — |
case-12 | fail→pass | 13,231 | 3,502 | -74% | 1 | 1 | 0% | 2,319 | 716 | -69% | 0 | 0 | — |
case-13 | pass→pass | 10,462 | 1,787 | -83% | 1 | 1 | 0% | 1,755 | 460 | -74% | 0 | 0 | — |
case-14 | fail→pass | 12,811 | 1,995 | -84% | 1 | 1 | 0% | 2,393 | 489 | -80% | 0 | 0 | — |
case-16 | fail→pass | 9,697 | 2,300 | -76% | 1 | 1 | 0% | 1,593 | 554 | -65% | 0 | 0 | — |
case-17 | fail→fail | 6,344 | 1,772 | -72% | 1 | 1 | 0% | 985 | 440 | -55% | 0 | 0 | — |
case-18 | fail→pass | 11,443 | 2,270 | -80% | 1 | 1 | 0% | 2,170 | 545 | -75% | 0 | 0 | — |
case-19 | fail→pass | 12,610 | 2,153 | -83% | 1 | 1 | 0% | 2,332 | 566 | -76% | 0 | 0 | — |
case-20 | fail→pass | 7,351 | 2,095 | -72% | 1 | 1 | 0% | 1,297 | 525 | -60% | 0 | 0 | — |
case-21 | pass→pass | 6,574 | 2,131 | -68% | 1 | 1 | 0% | 1,091 | 461 | -58% | 0 | 0 | — |
case-22 | fail→pass | 3,212 | 1,970 | -39% | 1 | 1 | 0% | 547 | 489 | -11% | 0 | 0 | — |
case-23 | pass→pass | 7,533 | 1,746 | -77% | 1 | 1 | 0% | 1,244 | 402 | -68% | 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. 23 cases were attempted. The headline lift of +57 percentage points is the difference between those two pass rates over the 23 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.