My Cursor AI workflow

Hardware	Any machine that runs VS Code comfortably
Software	Cursor (2024 builds, as originally written; refresh for current versions pending)
Date	2024-09-01

If you’re like me, you’ve probably found yourself diving deep into the world of Cursor AI. The idea of turning a simple concept into functional code within minutes is nothing short of revolutionary. But before you get ahead of yourself, it’s crucial to understand the best practices that can help you harness the full potential of Cursor AI.

In this post, I’ll share seven essential tips that have transformed the way I use Cursor AI. These insights come from both personal experience and expert advice from seasoned developers. Whether you’re a seasoned coder or just starting out, these strategies will save you hours of time and countless headaches.

Plan Before You Code
Leverage Cursor Directory for Better Prompts
Tag Relevant Documentation
Consult Other AI Models When Stuck
Use AI to Explain and Learn Code
Duplicate and Modify Existing Functionality
Utilize Existing Code Templates
Conclusion

Plan Before You Code

Measure twice, cut once. This age-old adage holds true even in the realm of AI-assisted coding. Before jumping into Cursor AI, take the time to plan your project thoroughly.

I can’t stress this enough: planning is crucial. Before jumping into Cursor AI, take some time to sketch out what you want to build. Whether it’s on paper, Paint, or Figma, having a visual reference makes a huge difference.

For example, when I was working on a web application recently, I drew a simple wireframe on my iPad. This helped me visualize the layout and functionality before I even started coding. Trust me, spending some time planning can save you hours later on.

Sketch Your Ideas: Use tools like Figma or even a simple pen and paper to outline your project.
Visualize with V0 (By Vercel): Platforms like V0 can help you create wireframes quickly. For example, if you’re building a marketplace for bicycle sellers, V0 can generate a clean-looking interface based on your prompts.

Why This Matters: Providing Cursor AI with a clear plan and visual references gives it more context, resulting in more accurate and efficient code generation.

Leverage Cursor Directory for Better Prompts

Did you know there’s a resource called Cursor Directory that can significantly improve your prompts?

Copy Expert Prompts: Find prompts tailored to the technologies you’re using, such as Next.js, React, or Python.
Set Up Cursor Rules: By creating a cursor-rules file in the root of your project and pasting the appropriate prompt, you guide Cursor AI to align with your project’s specific needs.

Example Prompt:

You are an expert in TypeScript, Node.js, Next.js, React, and Tailwind CSS. Follow best practices and write clean, maintainable code.

Benefits:

Improved Accuracy: Cursor AI will generate code that’s consistent with the latest best practices.
Time-Saving: Reduces the need for multiple iterations and corrections.

Tag Relevant Documentation

Context is king when it comes to AI models.

Add Docs to Cursor: Use the Add Docs feature in Cursor to import relevant documentation, such as the Next.js documentation or Supabase docs.
Stay Updated: This ensures Cursor AI references the most current information, helping you avoid outdated methods or deprecated functions.

How-To:

Go to the Composer view in Cursor.
Click on Add > Docs.
Paste the URL of the documentation you want to add.
Name it appropriately for easy reference.

Outcome: With tagged documentation, Cursor AI can provide more accurate solutions and even help troubleshoot issues using the latest guidelines.

Consult Other AI Models When Stuck

Even AI has its limitations. If you find yourself stuck, don’t hesitate to consult other AI models like Claude or even ChatGPT.

Steps:

Copy the Problem: Include the code that’s causing issues and the solutions you’ve already tried.
Provide Context: Explain what you’re trying to achieve and why the previous attempts didn’t work.
Ask for Help: Paste this into another AI model and see what solutions it offers.

Why This Works:

Different AI models may have varying datasets and can provide alternative solutions.
Including previous attempts helps the AI avoid suggesting the same ineffective solutions.

Use AI to Explain and Learn Code

AI isn’t just for generating code; it’s also a fantastic learning tool.

Explain Code Snippets: Ask Cursor AI to explain complex pieces of code as if you were a beginner.
Dive Deeper: If something doesn’t make sense, ask follow-up questions.

Example:

Explain this code to me like I'm a beginner. I want to understand the flow, logic, and overall how things work.

Benefits:

Accelerated Learning: Understand new technologies or frameworks quickly.
Better Documentation: Use AI-generated explanations to create documentation for your projects.

Duplicate and Modify Existing Functionality

Reusability is a cornerstone of efficient coding.

Copy Existing Code: If you have functionality that works well, copy it as a starting point for new features.
Specify Changes: Tell Cursor AI exactly what modifications you need.

Tips:

Provide Context: Mention where the code is from and what it currently does.
Detail the Changes: Be specific about what you want to alter to avoid unnecessary iterations.

Utilize Existing Code Templates

Starting from scratch isn’t always the best approach.

Use Starter Kits: Templates often include essential features like authentication, database integration, and basic UI components.
Customize: Build on top of these templates to suit your specific needs.

Resources:

GitHub: Search for reputable starter templates related to your tech stack.
Community Contributions: Many developers share their own templates, which can be a goldmine for getting started quickly.

Advantages:

Saves Time: Skip the boilerplate and focus on unique features.
Reduces Errors: Templates are often vetted by the community, minimizing bugs.

Conclusion

Harnessing the power of Cursor AI isn’t just about jumping in and coding; it’s about setting yourself up for success through careful planning and leveraging all the tools at your disposal.

Remember, the journey to mastering AI-assisted coding is a marathon, not a sprint. Embrace the learning process, and don’t be afraid to explore new methods and resources. As someone who’s navigated massive infrastructure projects and complex engineering challenges, I can attest that the time you invest in planning and learning today will pay dividends in the future.

By implementing these tips, I’ve been able to get the most out of Cursor AI and significantly improve my coding workflow. I hope these insights help you on your journey as well.

If this tutorial helped you out, I’d really appreciate it if you signed up for my newsletter.

Feel free to share your thoughts or ask questions in the comments below. You can also reach out to me on X (Formerly Twitter) @MeshRHM

Table of Contents