interface DataRecord {
  id: string | number;
  createdAt: Date;
}

interface Repository<TModel extends DataRecord, TKey extends keyof TModel> {
  findById(id: TModel[TKey]): TModel | undefined;
  findAll(filter?: Partial<TModel>): TModel[];
  save(entity: Omit<TModel, 'id' | 'createdAt'>): TModel;
}

function createPipeline<TInput, TOutput>(
  transform: (input: TInput) => TOutput,
  validate: (output: TOutput) => boolean
) {
  return (raw: TInput): TOutput | null => {
    const processed = transform(raw);
    return validate(processed) ? processed : null;
  };
}

const userPipeline = createPipeline(
  (raw: unknown) => raw as { name: string; email: string },
  (user) => typeof user.name === 'string' && user.email.includes('@')
);

for Dynamic Return Shapes

When a generic function's return type depends on the input structure, conditional types provide precise control without overloading.

type ApiResponse<TData, TError = Error> = 
  TData extends null ? { success: false; error: TError } : { success: true; data: TData };

function fetchResource<T>(endpoint: string): Promise<ApiResponse<T>> {
  return new Promise((resolve) => {
    // Simulated network layer
    const hasError = Math.random() > 0.8;
    if (hasError) {
      resolve({ success: false, error: new Error('Network failure') });
    } else {
      resolve({ success: true, data: {} as T });
    }
  });
}

Rationale: ApiResponse branches based on whether TData is null. This pattern replaces union types or optional chaining with a discriminated structure that the compiler can narrow automatically. Consumers get precise autocomplete and exhaustive type checking without manual type guards.

Step 4: Compose with Mapped Types for Configuration Objects

Generic configuration layers benefit from mapped types that preserve key relationships while allowing partial overrides.

type ConfigTemplate<T extends Record<string, unknown>> = {
  [K in keyof T]: {
    defaultValue: T[K];
    required: boolean;
    description: string;
  };
};

const dbConfig: ConfigTemplate<{ host: string; port: number; ssl: boolean }> = {
  host: { defaultValue: 'localhost', required: true, description: 'Database hostname' },
  port: { defaultValue: 5432, required: true, description: 'Connection port' },
  ssl: { defaultValue: false, required: false, description: 'Enable TLS' }
};

Rationale: Mapped generics enforce structural consistency across configuration objects. Adding a new config key automatically requires its metadata definition. This prevents drift between runtime expectations and type definitions.

Pitfall Guide

Generic abstractions introduce complexity that can backfire if misapplied. The following pitfalls represent the most common production failures, along with proven fixes.

1. The any Leak

Explanation: Developers insert any inside generic functions to bypass compiler errors, silently breaking type flow. The generic parameter becomes decorative rather than functional. Fix: Replace any with unknown and apply type guards or constraints. If a value must bypass strict checking, isolate it behind a clearly marked unsafe boundary and document the contract.

2. Over-Constraining with object

Explanation: Using T extends object blocks primitives, arrays, and functions. Many utilities legitimately need to accept strings, numbers, or booleans. Fix: Use T extends Record<string, unknown> for object shapes, or remove constraints entirely if the function operates on any type. Prefer specific interfaces over broad base types.

3. Forcing Explicit Type Arguments

Explanation: Writing createPipeline<User, UserDTO>(...) when TypeScript can infer both parameters from the callbacks. This adds noise and breaks when callback signatures change. Fix: Rely on inference. Only specify type parameters when returning a type that cannot be derived from arguments, or when working with higher-order generics.

4. Ignoring Variance Rules

Explanation: Assuming generic types are freely assignable in both directions. TypeScript enforces bivariance in function parameters by default, which can allow unsafe assignments. Fix: Enable strictFunctionTypes in tsconfig.json. Use readonly modifiers to signal covariance. Understand that Repository<User> is not assignable to Repository<DataRecord> unless explicitly designed for substitution.

5. Recursive Generic Loops

Explanation: Deeply nested or recursive generic types cause the compiler to hit recursion limits, resulting in Type instantiation is excessively deep and possibly infinite errors. Fix: Break complex types into smaller, composable pieces. Use infer sparingly and prefer conditional branching over deep recursion. Cache intermediate types using type aliases.

6. Missing Default Type Parameters

Explanation: APIs that require explicit type arguments for every call reduce adoption and increase boilerplate. Fix: Provide sensible defaults like T = unknown or TError = Error. Defaults allow consumers to opt-in to strict typing only when necessary.

7. Expecting Generics at Runtime

Explanation: Treating generic parameters as available during execution. TypeScript erases all type information during compilation; T does not exist in JavaScript. Fix: Pair generics with runtime validation libraries (Zod, Valibot, io-ts) when parsing external data. Use generics for compile-time contracts and runtime validators for actual data shape verification.

Production Bundle

Action Checklist

• Enable strict: true in tsconfig.json to activate strictNullChecks, strictFunctionTypes, and noImplicitAny
• Replace any usages with unknown and apply type guards or constraints before generic integration
• Define base interfaces for domain models and use extends to narrow generic parameters
• Rely on type inference for function arguments; specify type parameters only for return shapes or higher-order composition
• Add runtime validation layers for external data, keeping generics for internal type flow
• Document generic contracts with JSDoc, explicitly stating constraints, defaults, and expected behavior
• Monitor compiler performance; break recursive or deeply nested generics into smaller type aliases if build times degrade

Decision Matrix

Scenario	Recommended Approach	Why	Cost Impact
Simple data transformation	Generic function with inferred parameters	Minimal boilerplate, compiler handles type flow	Low (setup time only)
Repository / Data access layer	Constrained generic interface with TModel extends BaseRecord	Enforces consistent CRUD contracts across entities	Medium (initial interface design)
Event system / Pub-Sub	Generic event emitter with discriminated payload types	Type-safe subscription and emission without union sprawl	Medium (type definition overhead)
Plugin / Extension architecture	Generic factory with conditional return types	Allows plugins to declare their own config and output shapes	High (requires careful variance management)

Configuration Template

{
  "compilerOptions": {
    "target": "ES2022",
    "module": "ESNext",
    "strict": true,
    "noImplicitAny": true,
    "strictNullChecks": true,
    "strictFunctionTypes": true,
    "exactOptionalPropertyTypes": true,
    "noUncheckedIndexedAccess": true,
    "skipLibCheck": true,
    "forceConsistentCasingInFileNames": true,
    "declaration": true,
    "declarationMap": true,
    "sourceMap": true
  },
  "include": ["src/**/*.ts"],
  "exclude": ["node_modules", "dist", "**/*.test.ts"]
}

// src/generics/pipeline.ts
export type Transformer<TIn, TOut> = (input: TIn) => TOut;
export type Validator<T> = (value: T) => boolean;

export function buildPipeline<TIn, TOut>(
  transform: Transformer<TIn, TOut>,
  validate: Validator<TOut>
): (raw: TIn) => TOut | null {
  return (raw) => {
    const result = transform(raw);
    return validate(result) ? result : null;
  };
}

Quick Start Guide

1. Initialize strict mode: Create a tsconfig.json with strict: true and verify the compiler rejects implicit any and unsafe null access.
2. Define a base contract: Create an interface like DataRecord with required fields (id, createdAt) that all domain models will extend.
3. Build a constrained utility: Implement a generic function or class that accepts T extends DataRecord. Test it with two different domain types to verify inference and constraint enforcement.
4. Add runtime validation: Integrate a lightweight validator (e.g., Zod) for external data entry points. Use generics for internal type flow and validators for boundary enforcement.
5. Audit and refactor: Search for any and duplicated logic. Replace with generic abstractions, ensuring each replacement passes strict type checking and maintains existing behavior.