Install any skill in seconds. Free to start, no credit card required.
Get Started Free →TypeScript development best practices, code quality tools, and documentation conventions. Use when writing, editing, or reviewing .ts or .tsx files, when setting up a TypeScript project, when the user asks about TypeScript patterns/conventions, or when type safety guidance is needed.
.claude/skills/aiskillstore-typescript-dev/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 151% | 0% |
| case-17 | ✓→✗ | ▼ Worse | 16% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 36% | 0% |
This skill supports TypeScript project development.
pnpm as package managernpm or yarnstrict: true required?. and nullish coalescing ??require()any type in production codebash# Format code pnpm run format # Run linter pnpm run lint # Type check pnpm tsc --noEmit # Run tests with coverage pnpm test -- --coverage
Check these during code review:
any type usage?.) and Nullish coalescing (??) usedrequire())typescript// Good: Clear error types class ValidationError extends Error { constructor(message: string, public field: string) { super(message); this.name = 'ValidationError'; } } // Good: Result type pattern type Result<T, E = Error> = | { success: true; data: T } | { success: false; error: E };
typescript// Good: With error handling async function fetchUserData(id: string): Promise<Result<UserData>> { try { const response = await fetch(`/api/users/${id}`); const data = await response.json(); return { success: true, data }; } catch (error) { return { success: false, error: error as Error }; } }
typescript// Good: Custom type guard function isUserProfile(value: unknown): value is UserProfile { return ( typeof value === 'object' && value !== null && 'id' in value && 'username' in value ); }
React.memo for expensive componentsuseMemo / useCallback appropriatelyReact.lazy() for componentstypescript import type { UserProfile } from './types';
❌ Don't:
typescript// Using any type function process(data: any) { } // Implicit any function getValue(obj, key) { } // Excessive type assertions const user = data as User;
✅ Do:
typescript// Proper type definitions function process(data: UserData) { } // Explicit types function getValue<T>(obj: T, key: keyof T) { } // Use type guards if (isUser(data)) { // data is User here }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→pass | 24,285 | 12,468 | -49% | 1 | 1 | 0% | 2,830 | 3,435 | +21% | 0 | 0 | — |
case-01 | fail→fail | 18,086 | 21,051 | +16% | 1 | 1 | 0% | 3,831 | 4,207 | +10% | 0 | 0 | — |
case-03 | pass→pass | 8,252 | 11,018 | +34% | 1 | 1 | 0% | 1,516 | 2,062 | +36% | 0 | 0 | — |
case-04 | pass→pass | 13,896 | 11,072 | -20% | 1 | 1 | 0% | 1,614 | 1,982 | +23% | 0 | 0 | — |
case-05 | fail→fail | 12,799 | 15,748 | +23% | 1 | 1 | 0% | 2,728 | 2,993 | +10% | 0 | 0 | — |
case-06 | pass→pass | 12,239 | 11,768 | -4% | 1 | 1 | 0% | 2,354 | 3,039 | +29% | 0 | 0 | — |
case-07 | pass→pass | 8,482 | 9,195 | +8% | 1 | 1 | 0% | 630 | 1,601 | +154% | 0 | 0 | — |
case-08 | pass→pass | 14,421 | 11,523 | -20% | 1 | 1 | 0% | 1,607 | 1,988 | +24% | 0 | 0 | — |
case-09 | pass→pass | 10,009 | 5,302 | -47% | 1 | 1 | 0% | 816 | 1,897 | +132% | 0 | 0 | — |
case-10 | pass→pass | 9,498 | 6,570 | -31% | 1 | 1 | 0% | 777 | 2,098 | +170% | 0 | 0 | — |
case-11 | pass→pass | 11,299 | 9,165 | -19% | 1 | 1 | 0% | 1,063 | 1,866 | +76% | 0 | 0 | — |
case-12 | pass→pass | 32,120 | 30,185 | -6% | 1 | 1 | 0% | 3,104 | 3,651 | +18% | 0 | 0 | — |
case-13 | pass→pass | 9,010 | 6,622 | -27% | 1 | 1 | 0% | 672 | 2,032 | +202% | 0 | 0 | — |
case-14 | pass→pass | 3,854 | 8,637 | +124% | 1 | 1 | 0% | 644 | 1,578 | +145% | 0 | 0 | — |
case-15 | pass→pass | 7,082 | 7,444 | +5% | 1 | 1 | 0% | 348 | 1,352 | +289% | 0 | 0 | — |
case-16 | fail→pass | 16,308 | 8,783 | -46% | 1 | 1 | 0% | 2,257 | 2,618 | +16% | 0 | 0 | — |
case-17 | pass→fail | 13,958 | 11,102 | -20% | 1 | 1 | 0% | 1,736 | 2,014 | +16% | 0 | 0 | — |
case-18 | fail→pass | 2,947 | 7,350 | +149% | 1 | 1 | 0% | 505 | 1,269 | +151% | 0 | 0 | — |
case-19 | pass→pass | 15,248 | 9,141 | -40% | 1 | 1 | 0% | 1,931 | 2,517 | +30% | 0 | 0 | — |
case-20 | pass→pass | 9,438 | 8,626 | -9% | 1 | 1 | 0% | 1,821 | 2,458 | +35% | 0 | 0 | — |
case-21 | pass→pass | 3,695 | 3,838 | +4% | 1 | 1 | 0% | 584 | 1,592 | +173% | 0 | 0 | — |
case-22 | pass→pass | 7,866 | 9,396 | +19% | 1 | 1 | 0% | 444 | 1,638 | +269% | 0 | 0 | — |
case-23 | pass→pass | 3,336 | 3,605 | +8% | 1 | 1 | 0% | 561 | 1,515 | +170% | 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 +9 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.