Built and Open-Sourced 3 Mini Tools with Vibe Coding

in #blog • 7 days ago

Recently, I used ChatGPT-4o and o4-mini to quickly build and open-source three handy tools. The motivation was simple—every time I needed to convert YAML/JSON or do Base64 encoding, I found myself Googling for some online tools. Then I thought, why not just build my own?

At the same time, I’ve been learning frontend development with React, so it felt like the perfect opportunity to put that into practice.

So I basically played the role of a product manager—gave my requirements to ChatGPT, and it quickly set up the entire framework for me. That included project scaffolding, unit testing, ESLint (which I later replaced with Prettier), README docs, and even CI/CD with GitHub Actions for automated testing and deployment.

All three tools are deployed on GitHub Pages—fully open-source, with tests and documentation. They’re small, but complete. And since they’re static web tools, there’s no need for a backend server at all.

YAML/JSON Converter


A simple and clean tool for converting between YAML and JSON formats. It supports both prettified and minified output.

[caption id="attachment_156997" align="alignnone" width="1321"]Yaml Json Converter (Vibe Coding), React Js Web App Yaml Json Converter (Vibe Coding), React Js Web App[/caption]

Github Repo | Github Pages Tool

Base64 Encoder/Decoder


Supports both plain text and file encoding/decoding. You can drop in a file to get a Base64 string, or paste a Base64 string to recover the original file.

[caption id="attachment_156998" align="alignnone" width="1236"]Base 64 Converter (Vibe Coding), React Js Web App Base 64 Converter (Vibe Coding), React Js Web App[/caption]

Github Repo | Github Pages Tool

ROT47 Obfuscation Tool


ROT47 is a reversible character substitution cipher—just like XOR, applying it twice gives you back the original text. Fun for light obfuscation.

[caption id="attachment_156999" align="alignnone" width="900"]ROT47 Ciper Converter (Vibe Coding), React Js Web App ROT47 Ciper Converter (Vibe Coding), React Js Web App[/caption]

Github Repo | Github Pages Tool

The Barrier to Entry for Programming Is Dropping


AI is evolving at an insane pace. It writes better code now, and with much higher accuracy. Even if there are bugs, you can just copy-paste the error into ChatGPT and get a fix. In many cases, you no longer need to know how to write code line by line—just understand the requirement and know how to communicate clearly with AI.

For me, starting a new project used to be the hardest part—it's often the biggest source of procrastination. But that’s exactly what ChatGPT excels at: bootstrapping boilerplate code. It significantly boosts productivity by removing the friction from getting started.

I now rely on AI daily. It’s become a habit. I hardly Google things anymore—if I run into a problem, I just ask ChatGPT.

Vibe Coding: Letting the AI Handle the Details


"Vibe Coding" is a hot trend this year. The idea is: describe what you want in natural language, and let the AI write the code. It’s an AI-assisted coding practice where large language models (LLMs) take care of the implementation, while you guide the direction.

Andrej Karpathy (2025)

TLDR; Vibe coding is the emerging practice of building software by describing what you want in plain language - and letting AI handle the code and test.

I noticed I’m increasingly coding in this way—if there's a bug, I throw the stack trace into ChatGPT. While this is super efficient, it does take away some of the thinking process and the fun of debugging.

There are now tools built specifically for Vibe Coding—like Cursor, or AI Agents in VSCode. Just tell the agent what you want, and it’ll scaffold your project or write the logic for you. Your role shifts to reviewing and approving its changes—or asking it to tweak things further.

Isn’t that basically what a product manager does?

Entry-Level Developers Might Be at Risk


The most important thing now is to know how to use AI tools to improve your productivity. Just following tutorials and copying code is no longer enough.

You need to think more, design more, and learn how to work with AI, not against it.

AI can be a superpower—but only if you learn how to wield it. While the definition of “coding” may be changing, the essence of solving problems remains the same.

What is more important for Software Engineer? Probably creativity, execution, debugging and solving problems and knowing how to talk to AI (asking AI to do the work).

ChatGPT Use Cases

ChatGPT - AGI (Artificial General Intelligence)

Grok 3 AI

ChatGPT Fails to

image.png

Steem to the Moon🚀!

#vibecoding #ai #coding #development #softwareengineer
7 days ago in #blog by
$23.45
  • Pending payout amount: $23.45
  • Breakdown:
        0.00 SBD,
        75.18 STEEM,
        75.18 SP
  • Payout in 5 hours
142 votes
Reply 0