Vibe Coding was pretty revolutionary by Codex
#SRG(Service Reliability Group) is a group that mainly provides cross-sectional support for the infrastructure of our media services, improving existing services, launching new ones, and contributing to OSS.
This article is an account of my experience with vibe coding using Codex.
Introduction
In recent years, development support tools that utilize AI have been attracting attention.
Among these, a new development style called "vibe coding" is gaining attention.
I've been using the Agent mode of Cursor, an AI code editor, in my work.
Recently, the OpenAI Codex became available to our team, so we tried out vibe coding using this Codex.
In this article, I will introduce what vibe coding is and my experience with vibe coding using Codex.
What is vibe coding?
First, let's talk about vibe coding.
Vibe coding is a programming paradigm proposed by computer scientist Andrej Karpathy.
In this approach, a programmer provides a few sentences of instructions (prompts) in natural language to a large-scale language model (LLM), which then generates the software.
In this approach, the programmer's role changes from manually writing code to guiding, testing, and refining the AI-generated source code.
According to IBM, the goal of vibe coding is to create an AI-powered development environment where AI makes real-time suggestions, automates tedious processes, and generates a standard codebase structure.
The idea is to "write the code first, refine it later," which is said to be in line with agile principles such as rapid prototyping and iterative development.
Vibe coding with Codex
This time, I used OpenAI's Codex to experience vibe coding.
Codex is a cloud-based software engineering agent developed by OpenAI.
They can perform many tasks in parallel, such as developing features, answering questions about the code base, fixing bugs, and proposing pull requests.
I don't think there's any clear difference between this and Cursor's Agent mode.
(Autonomously explores the codebase, plans, and modifies complex codebases)
By the way, I have the following Cursor Rule written in my code to ensure that I search through it properly.
I feel that the difference between having "Search related files if necessary" and not having it is quite large.
What was particularly impressive when actually using Codex was how easy it was to create a pull request (PR) with just one button, without having to set up any special instructions like "Rules" in Cursor.
Cursor will not create PRs using GitHub MCP unless you explicitly tell it to.
The changes we tried this time were modifications to the sitemap.xml of the English version of the portal site you are currently viewing.
All the code is written in Cursor Agent using mostly vibe coding.
The details are as follows, and it is so complex that without AI, implementation would be abandoned.
Instructions
It goes against the principles of vibe coding to take the time to read through the code output by vibe coding, but I did take a look at it this time while writing this article, and I think the code output is easy to read and good.
Although it's not stated in the instructions, it's nice that they also updated the README.md.
In the case of Cursor, it is often not written (but if you write the Cursor Rule mentioned above, it will be done).
Conclusion
I personally found vibe coding with Codex to be "pretty good" and a very innovative experience.
With AI taking on most of the coding, developers could be freed up to focus on higher-level design and problem-solving.
Of course, reviewing and testing AI-generated code is essential, but we felt that vibe coding would be a powerful tool in the early stages of development and prototyping.
I hope that AI development support tools like Codex will continue to evolve and that vibe coding will become a more common development style in the future.
SRG is looking for people to work with us.
If you're interested, please contact us here.