👉 Spoiler Alert - my OSS Project for Github Repositories cleanup

Have you ever heard the phrase “Clean your room first” before starting something big?

That’s exactly how I kicked off this blog and my AI learning journey — by cleaning up my cluttered GitHub profile.

Problem: 325 Forked Repos

At some point, I got obsessed with forking every interesting repo I came across.
Eventually, it became a mess — 325 repos, most untouched, all clogging my dashboard.

Manually deleting them? Absolutely not.

GitHub profile with cluttered forks

My Rule: Solve My Problems With Code

I believe in learning by solving your own problems — and this was the perfect case.
So I set out to write a small Python tool that would:

  • List all my repositories
  • Show basic info + readme preview
  • Ask me which ones to delete
  • Allow marking protected repos (via a config file)
  • Batch delete confirmed ones
  • Support safer actions: archive or make private

Building It: On an Old Tablet, With AI Help

I built everything on a Surface tablet running Windows 10 — far from ideal, but it worked.

I used:

  • Visual Studio Code
  • GitHub Copilot (free tier)
  • ChatGPT to structure and debug
  • A lot of coffee

Setting up Copilot in VS Code

Bugs? Of Course There Were Bugs

The tool didn’t work at first — common story.

I hit problems like:

  • Missing GitHub username during API calls
  • Repos not deleting due to permission errors
  • Debugging token storage

Early error during repo deletion

But I fixed them iteratively. That’s what building looks like.

Iterating: Batch Delete & Safeguards

Once the basics worked, I needed batch deletion. So I added it.

Batch delete workflow

I also made sure the tool doesn’t delete:

  • Starred repos
  • Whitelisted names in a config
  • Anything without confirmation

And instead of just deleting, it now offers to:

  • Archive the repo
  • Make it private

Much safer.

Sharing It as Open Source

Eventually, I decided to open source it. Why?

Because:

  • It helped me
  • Others have the same problem
  • It’s a simple but empowering first project

👉 Check it out on my GitHub

What I Learned

  • Solve your own pain first
  • Start tiny — grow through usage
  • Don’t wait until you’re an expert
  • Ship something that helps you — and share it

What’s Next

This is just the first of many experiments.
Next up: building small automations and AI workflows to create alternative income streams — and documenting all of it right here.

The blog, like the projects, will evolve as I do.

Thanks for reading.