I Wrote My First Chrome Extension Without a Single Line of Code

A colleague recently asked me if I knew of a QA tool that could check whether a coordinate was within a polygon. I replied that I didn’t, but suggested we try to build a tool like that with ChatGPT.

In this blog post, I will detail how I built a Chrome extension using ChatGPT without writing a single line of code, only following instructions and copy-pasting, despite having no prior experience in creating a Chrome app.

👉 Read the full ChatGPT transcript here
👉 Install from the Chrome Web Store: Polygon Checker
👉 Source on GitHub: github.com/dannyshmueli/polygon-checker

The Idea

The goal was to build a Chrome extension that could take a list of coordinates describing a polygon and a single map coordinate as input, then output whether the single coordinate lies inside the polygon.

The Process

I began by explaining my idea to ChatGPT 3.5. The model understood the requirements and produced everything I needed:

1. A manifest.json file.
2. A popup.html UI.
3. A JavaScript file with the point-in-polygon logic.
4. Step-by-step instructions for loading and testing the extension in Chrome.

ChatGPT even helped me debug along the way – for example, fixing an issue where the result disappeared too quickly after display.

The first version’s UI wasn’t ideal (the textbox was tiny), so I asked ChatGPT to enlarge it and polish the styling. After a few iterations the popup looked and felt great.

I also noticed the generated manifest.json listed unnecessary permissions. After asking “Why do we need these?” ChatGPT agreed they could be removed and produced a leaner version:

Finally, ChatGPT wrote a perfectly-serviceable README.md, which I committed as-is.

Reflections on AI Collaboration

Working closely with ChatGPT offered a unique perspective on what AI collaboration can look like in product management and software development.

Accelerating development

The pace was remarkable. What would normally require hours of searching and trial-and-error took minutes with concise prompts and copy-paste.

Bridging the knowledge gap

I had never built a Chrome extension before, yet ChatGPT walked me through every step, from project structure to Chrome’s developer mode.

Enhancing creativity

Because ChatGPT handled the boilerplate, I could focus on the creative parts – UI tweaks, use-case brainstorming, and of course writing this post.

Potential challenges

Not everything was perfect – for example, the suggested SVG icon needed a redo. Human oversight still matters, and the model will only get better.

Conclusion

This experience felt like pairing with a tireless coworker, iterating together until we were both satisfied. It showed me how AI tools can empower anyone – regardless of background – to create and ship useful software.

As language models evolve, I can only imagine how they’ll continue to revolutionise software engineering and product development.