llm.md

5 ChatGPT Prompts to Accelerate Code Refactoring

PromptCraft Team
#programming #chatgpt #refactoring

Refactoring code is a vital but time-consuming part of software engineering. Modern LLMs are exceptional at identifying code smells, simplifying complex logic, and applying design patterns.

However, generic requests like “refactor this code” often lead to code that doesn’t compile, changes functionality, or doesn’t follow your team’s style guide.

To get clean, production-ready code on the first try, use these five highly specific prompt templates.


1. The “Big-O” Complexity Optimizer

Use this prompt when you have working code that is too slow, memory-intensive, or uses nested loops.

Act as a principal software performance engineer. Review the following code for time and space complexity. 

Optimize it to reduce its Big-O complexity (e.g., from O(n^2) to O(n)). Maintain exact functional parity. 

Explain the optimizations you made and detail the time/space complexity before and after.

<code>
[Insert code here]
</code>

2. The “Clean Code & Readability” Refactorer

Use this prompt to turn quick-and-dirty code into readable, maintainable, and self-documenting code.

Act as a senior software architect. Refactor this code to improve readability and maintainability. 

Follow these clean code guidelines:
- Rename obscure variables/functions to descriptive names.
- Break down monolithic functions into smaller, single-responsibility functions.
- Remove redundant comments, leaving only explanatory comments for non-obvious code.
- Ensure strict type safety.

<code>
[Insert code here]
</code>

3. The “Design Pattern” Integrator

Use this prompt when you need to restructure code to use a standard software design pattern (e.g., Factory, Singleton, Strategy).

Act as a software design pattern expert. Refactor the following code to implement the [Insert Pattern, e.g., Strategy] design pattern. 

Ensure the implementation is clean, decoupled, and makes the codebase easier to extend in the future. Explain how the pattern is applied.

<code>
[Insert code here]
</code>

4. The Unit Test Writer

Refactored code must be tested. Use this prompt to generate test suites covering happy paths, edge cases, and error modes.

Act as a senior QA engineer. Write a comprehensive unit test suite for the following code using [Insert framework, e.g., Jest/Vitest/PyTest].

Ensure you cover:
- Typical inputs (happy paths)
- Boundary values and edge cases (empty strings, null, negative numbers)
- Expected error conditions and exceptions

<code>
[Insert code here]
</code>

5. Legacy Code Modernizer

Use this prompt when you are updating older codebases (e.g., converting ES5 JavaScript to ES6+ TypeScript or Python 2 to Python 3).

Act as a legacy code migration expert. Modernize the following [Insert language/version, e.g., ES5 JavaScript] code to [Insert target, e.g., modern TypeScript].

Apply modern language features (arrow functions, async/await, destructured objects, typed arguments) and clean up deprecated methods.

<code>
[Insert code here]
</code>

Tip: Refining Your Prompts Instantly

For the best results, paste your refactored code and the optimized output side-by-side in your editor to verify functionality before committing.

If you want to create your own custom coding assistant prompts, try running them through PromptCraft on our homepage to automatically enhance them with target contexts and negative constraints.

Refine Your AI Prompts Automatically

Put the prompt engineering concepts in this guide to work. Use PromptCraft to instantly rewrite, structure, and optimize your prompts.