Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Whole-repo audit for over-engineering. Like ponytail-review, but scans the entire codebase instead of a diff: a ranked list of what to delete, simplify, or replace with stdlib/native equivalents. Use when the user says \"audit this codebase\", \"audit for over-engineering\", \"what can I delete from this repo\", \"find bloat\", \"ponytail-audit\", or \"/ponytail-audit\". One-shot report, does not
.claude/skills/kunanonj-cursor-plugin-ponytail-ponytail-audit/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 48% | 0% |
ponytail-review, repo-wide. Scan the whole tree instead of a diff. Rank findings biggest cut first.
Same as ponytail-review:
delete: dead code, unused flexibility, speculative feature. Replacement: nothing.stdlib: hand-rolled thing the standard library ships. Name the function.native: dependency or code doing what the platform already does. Name the feature.yagni: abstraction with one implementation, config nobody sets, layer with one caller.shrink: same logic, fewer lines. Show the shorter form.Deps the stdlib or platform already ships, single-implementation interfaces, factories with one product, wrappers that only delegate, files exporting one thing, dead flags and config, hand-rolled stdlib.
One line per finding, ranked: <tag> <what to cut>. <replacement>. [path]. End with net: -<N> lines, -<M> deps possible. Nothing to cut: Lean already. Ship.
Complexity only, correctness bugs, security holes, and performance go to a normal review pass. Lists findings, applies nothing. One-shot. "stop ponytail-audit" or "normal mode" to revert.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→pass | 4,828 | 2,319 | -52% | 1 | 1 | 0% | 861 | 749 | -13% | 0 | 0 | — |
case-08 | fail→fail | 11,522 | 5,917 | -49% | 1 | 1 | 0% | 2,062 | 1,463 | -29% | 0 | 0 | — |
case-14 | pass→pass | 7,831 | 2,572 | -67% | 1 | 1 | 0% | 1,534 | 821 | -46% | 0 | 0 | — |
case-01 | fail→fail | 7,042 | 3,737 | -47% | 1 | 1 | 0% | 1,412 | 420 | -70% | 0 | 0 | — |
case-02 | fail→fail | 3,197 | 2,902 | -9% | 1 | 1 | 0% | 290 | 567 | +96% | 0 | 0 | — |
case-13 | fail→pass | 9,333 | 1,645 | -82% | 1 | 1 | 0% | 1,813 | 580 | -68% | 0 | 0 | — |
case-03 | fail→fail | 8,821 | 2,486 | -72% | 1 | 1 | 0% | 1,565 | 416 | -73% | 0 | 0 | — |
case-04 | fail→fail | 9,342 | 10,889 | +17% | 1 | 1 | 0% | 1,112 | 1,563 | +41% | 0 | 0 | — |
case-05 | pass→pass | 12,520 | 17,796 | +42% | 1 | 1 | 0% | 2,129 | 3,452 | +62% | 0 | 0 | — |
case-06 | fail→fail | 4,106 | 18,948 | +361% | 1 | 1 | 0% | 200 | 3,801 | +1801% | 0 | 0 | — |
case-07 | fail→fail | 2,219 | 2,053 | -7% | 1 | 1 | 0% | 281 | 651 | +132% | 0 | 0 | — |
case-09 | fail→pass | 9,009 | 2,241 | -75% | 1 | 1 | 0% | 1,581 | 640 | -60% | 0 | 0 | — |
case-10 | fail→pass | 2,182 | 2,585 | +18% | 1 | 1 | 0% | 342 | 718 | +110% | 0 | 0 | — |
case-11 | fail→pass | 3,170 | 3,143 | -1% | 1 | 1 | 0% | 549 | 814 | +48% | 0 | 0 | — |
case-12 | fail→pass | 2,928 | 2,799 | -4% | 1 | 1 | 0% | 587 | 812 | +38% | 0 | 0 | — |
case-16 | fail→pass | 7,727 | 2,449 | -68% | 1 | 1 | 0% | 1,405 | 757 | -46% | 0 | 0 | — |
case-17 | fail→pass | 5,161 | 3,604 | -30% | 1 | 1 | 0% | 1,060 | 1,021 | -4% | 0 | 0 | — |
case-18 | fail→pass | 13,804 | 5,857 | -58% | 1 | 1 | 0% | 2,763 | 1,409 | -49% | 0 | 0 | — |
case-19 | fail→fail | 14,695 | 4,090 | -72% | 1 | 1 | 0% | 2,873 | 500 | -83% | 0 | 0 | — |
case-20 | fail→fail | 12,904 | 3,486 | -73% | 1 | 1 | 0% | 2,285 | 365 | -84% | 0 | 0 | — |
case-21 | fail→pass | 4,889 | 3,340 | -32% | 1 | 1 | 0% | 1,002 | 981 | -2% | 0 | 0 | — |
case-22 | fail→pass | 10,223 | 3,697 | -64% | 1 | 1 | 0% | 1,771 | 976 | -45% | 0 | 0 | — |
case-23 | fail→pass | 12,338 | 6,112 | -50% | 1 | 1 | 0% | 2,459 | 1,648 | -33% | 0 | 0 | — |
case-24 | pass→pass | 10,846 | 6,719 | -38% | 1 | 1 | 0% | 2,141 | 1,276 | -40% | 0 | 0 | — |
case-25 | fail→fail | 13,502 | 6,441 | -52% | 1 | 1 | 0% | 2,290 | 1,534 | -33% | 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. 25 cases were attempted, and 19 counted toward the lift figure. The other 6 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 +48 percentage points is the difference between those two pass rates over the 19 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.