Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Post-AI-generation cleanup — remove single-use helpers, over-abstractions, generic naming, and unnecessary configurability. Use after dev-autopilot or any LLM-generated code burst.
.claude/skills/evolution-foundation-dev-ai-slop-cleaner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -41% | 0% |
Derived from oh-my-claudecode (MIT, Yeachan Heo). Adapted for the EvoNexus Engineering Layer.
Cleanup pass for code generated by AI. Removes the typical bloat: single-use helpers, over-abstractions, generic names like processData(), unnecessary configurability that wasn't asked for.
dev-autopilot finishes a feature@bolt-executor produces a multi-file change@apex-architect first)typescript// SLOP function calculateSum(a: number, b: number): number { return a + b; } const total = calculateSum(price, tax); // CLEANED const total = price + tax;
typescript// SLOP class TimeoutConfigManager { private static instance: TimeoutConfigManager; static getInstance() { ... } setTimeout(ms: number) { ... } getTimeout() { ... } } // CLEANED const TIMEOUT_MS = 5000;
typescript// SLOP function processData(data: any): any { ... } // CLEANED function parseCSVRow(row: string): User { ... }
typescript// SLOP function fetchUser(id: string, options?: { retries?: number, backoff?: number, timeout?: number, cache?: boolean }) { ... } // CLEANED (if retries/etc. weren't asked for) function fetchUser(id: string) { ... }
Remove or convert to actual issues.
@bolt-executor to apply (or @zen-simplifier for batch cleanup)dev-verify to confirm behavior is unchangedworkspace/development/research/[C]slop-cleanup-{component}-{date}.md@zen-simplifier (the agent that does the actual cleanup)@bolt-executor (when fixes are simple)dev-verify (to confirm zero behavior change)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 6,296 | 3,043 | -52% | 1 | 1 | 0% | 981 | 1,030 | +5% | 0 | 0 | — |
case-01 | fail→fail | 10,206 | 4,847 | -53% | 1 | 1 | 0% | 1,031 | 899 | -13% | 0 | 0 | — |
case-03 | fail→fail | 4,880 | 4,573 | -6% | 1 | 1 | 0% | 800 | 907 | +13% | 0 | 0 | — |
case-04 | pass→pass | 13,728 | 4,211 | -69% | 1 | 1 | 0% | 2,042 | 1,326 | -35% | 0 | 0 | — |
case-05 | fail→pass | 12,803 | 5,061 | -60% | 1 | 1 | 0% | 1,912 | 1,496 | -22% | 0 | 0 | — |
case-06 | pass→pass | 14,053 | 4,422 | -69% | 1 | 1 | 0% | 2,019 | 1,419 | -30% | 0 | 0 | — |
case-07 | pass→pass | 4,449 | 3,503 | -21% | 1 | 1 | 0% | 748 | 1,251 | +67% | 0 | 0 | — |
case-08 | pass→pass | 8,602 | 5,668 | -34% | 1 | 1 | 0% | 1,469 | 1,661 | +13% | 0 | 0 | — |
case-09 | fail→fail | 8,287 | 3,233 | -61% | 1 | 1 | 0% | 1,464 | 1,251 | -15% | 0 | 0 | — |
case-10 | pass→pass | 10,278 | 3,982 | -61% | 1 | 1 | 0% | 1,686 | 1,425 | -15% | 0 | 0 | — |
case-11 | pass→pass | 8,457 | 3,662 | -57% | 1 | 1 | 0% | 1,336 | 1,248 | -7% | 0 | 0 | — |
case-12 | pass→pass | 11,918 | 4,851 | -59% | 1 | 1 | 0% | 1,888 | 1,493 | -21% | 0 | 0 | — |
case-13 | pass→pass | 11,031 | 3,778 | -66% | 1 | 1 | 0% | 1,688 | 1,314 | -22% | 0 | 0 | — |
case-14 | fail→pass | 7,497 | 2,277 | -70% | 1 | 1 | 0% | 1,301 | 1,056 | -19% | 0 | 0 | — |
case-15 | fail→pass | 9,105 | 2,971 | -67% | 1 | 1 | 0% | 1,430 | 1,160 | -19% | 0 | 0 | — |
case-16 | fail→pass | 5,322 | 2,119 | -60% | 1 | 1 | 0% | 812 | 1,030 | +27% | 0 | 0 | — |
case-17 | fail→pass | 9,754 | 1,905 | -80% | 1 | 1 | 0% | 1,550 | 912 | -41% | 0 | 0 | — |
case-18 | pass→pass | 8,923 | 3,855 | -57% | 1 | 1 | 0% | 1,548 | 1,340 | -13% | 0 | 0 | — |
case-19 | pass→pass | 9,375 | 5,030 | -46% | 1 | 1 | 0% | 1,560 | 1,553 | -0% | 0 | 0 | — |
case-20 | pass→pass | 8,423 | 4,084 | -52% | 1 | 1 | 0% | 1,391 | 1,352 | -3% | 0 | 0 | — |
case-21 | pass→pass | 13,714 | 5,174 | -62% | 1 | 1 | 0% | 1,449 | 1,531 | +6% | 0 | 0 | — |
case-22 | pass→pass | 12,751 | 6,023 | -53% | 1 | 1 | 0% | 1,920 | 1,617 | -16% | 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, and 20 counted toward the lift figure. The other 2 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 +23 percentage points is the difference between those two pass rates over the 20 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.