Cursor AI – How Tools Can Improve Developer Performance

Introduction

Recently, new tools have emerged on the market, offering exciting opportunities to enhance development performance. Curious about their potential, I decided to explore how they could improve my day-to-day workflow. Among these tools, one seems particularly interesting, Cursor AI, which stands out as a game-changer.

In simple terms, Cursor AI is an advanced code-editing and AI-assisted development tool designed to elevate the coding experience for developers. It boosts productivity, streamlines problem-solving, and makes tackling complex challenges more efficient. In this blog, I've compiled a list of benefits highlighting why Cursor AI is a must-try for any developer looking to optimise their workflow.

Overview

Cursor AI is a fast, intelligent code editor that empowers developers with AI-assisted capabilities. As developers, we often find ourselves copying and pasting snippets of code—whether to explain its functionality, pinpoint bugs, or simply ask why it's not working as expected. Cursor AI streamlines this process by combining a powerful code editor with an embedded chat feature, making writing, debugging, and understanding code easier.

Trusted by engineers at leading companies like Shopify, OpenAI, Samsung, and many more, Cursor AI's widespread adoption inspired me to dive in and explore its potential.

Since its inception, Cursor AI has quickly become a familiar tool for developers. This familiarity stems from its similarity to Visual Studio Code (VS Code), with an interface designed to make users feel right at home. Given that most developers today have either used or at least seen VS Code, this design choice ensures a seamless transition.

During the installation process, the first step prompts you to import key bindings, a feature from VS Code that feels like an obvious choice for many. However, support for other popular editors like Vim, Atom, and Sublime Text are also available. As shown in the screenshot below, you can configure the language for AI, enable your codebase, and even set up the option to open Cursor directly from the command line.

I opted for the VS Code setup, and the next step proved incredibly handy. Cursor AI allowed me to import extensions, settings, and key bindings directly from VS Code. This was a huge time-saver—my SQL Developer extension and all my shortcuts were imported flawlessly, making the transition effortless. There's also an option to start from scratch if that suits your needs better.

During the installation and configuration process, the final step is defining the privacy mode. As shown below, you have the option to select a private mode, ensuring that none of your code is stored by external companies.

Generate tables and relations

To test the performance of Cursor AI, I created a simple project folder to organise all my documents. As a first step, I decided to ask Cursor to generate tables for a master-detail structure with all the required elements. I started by taking a screenshot of a master-detail layout from my mock legacy app and used it as a reference to seek assistance.

The screenshot below illustrates my prompt and part of the results. I requested Cursor to create the master-detail tables based on the screenshot, and it performed incredibly well. Cursor not only built the tables but also generated sequences and indexes. Additionally, it provided options to ask questions, copy the generated code, and apply it directly.

I then asked Cursor to edit the script by adding a prefix to each column name, and it handled the task perfectly, showcasing its flexibility and precision.

In summary, Cursor AI exceeded my expectations. It successfully created a small table structure, eliminating the need to copy and paste scripts from a web browser.

Moving on to the next set of benefits, a standout feature is the ability to set the context for a requested prompt. As shown in the screenshot below, each prompt provides additional options such as:

• Model Selector
• Generate Response Based on Codebase
• Add Additional Context

Adding additional context is helpful. By using the context of a defined reference, Cursor can generate more accurate and relevant responses. This feature allows you to include files, folders, documents, and more, enhancing the precision of the output.

In the next section, I'll demonstrate more examples to show how powerful this capability can be.

Generate API and documentation

With the two tables created, I wanted to generate an API and accompanying documentation. This process is straightforward and only requires crafting a precise prompt. My prompt, as shown below, was designed to achieve this efficiently.

Below, I've included just a portion of the results. Everything was generated as I expected. If something isn't correct, we can easily refine the output by adding additional requests through the "Ask" option. Creating a file is incredibly simple—just two clicks: Apply and Create New File. As a result, the package body and specification are automatically created in my project folder.

Finally, with the tables and API completed, I needed to create standard documentation for the work I've done so far. I created a prompt to generate the documentation using all the files I had already created in the project folder.

Once again, I was not disappointed. The output included clear and solid documentation featuring descriptions of the tables and APIs. It also provided a brief overview of all columns and function parameters, which can be directly used as comments in the package body and specifications.

Unit Test

Testing is a critical part of every project, but creating unit tests can often feel tedious for developers. With Cursor AI, this process can be significantly accelerated. Cursor AI includes functionality to add documentation context, guiding the AI on which data sources to use for specific requests.

For testing purposes in my project, I added a link to the documentation for utPLSQL (an open-source testing framework for PL/SQL and SQL). Using this documentation as a reference, I asked Cursor to generate a few unit tests.

In the results, it automatically creates various unit tests—about 10 in total—that I can easily integrate into my testing process.

Other features

In addition to generating new code, Cursor AI offers a range of real-time coding assistance features. It provides autocomplete suggestions, helps identify bugs, explains what the code is doing, and much more. For instance, I used Cursor AI to correct an existing package. I selected a part of my code and asked it to make the corrections, and the results were impressive.

For each request, before applying any changes, the editor highlights what has been changed, removed, or added. I can choose whether to accept all changes or only a part of them. The final decision always rests with the developer.

When working on a large project with an extensive codebase, you can also ask Cursor to locate a specific function, and it will help you find and use it efficiently.

Summary

In my opinion, Cursor AI is an incredibly efficient tool for development work. While the free version has its limitations, the paid version is worth considering. One major benefit is that it eliminates the need to copy and paste code, streamlining the workflow significantly. Beyond code-focused prompts, you can ask for virtually anything you need assistance with.

Additionally, Cursor AI offers the flexibility to select AI models tailored to different types of prompts. There are also numerous options you can configure to suit your personal or project preferences. I highly recommend checking it out to experience its full potential.

For more information, check out our Development Services, and if you liked this blog, check out our other development blogs here.