How I Used OpenAI Codex to Migrate 200+ OpenProcessing Sketches to GitHub in 10 Minutes
In a world where digital creativity meets code, one of my biggest challenges has always been managing and organising dozens—sometimes hundreds—of small creative coding projects. I recently faced that very challenge with my collection of over 200 sketches hosted on OpenProcessing (learn more about these sketches on our Playground section). These were HTML and JavaScript sketches, some dating back years, scattered across my OpenProcessing profile.
Image: More than 200 clickable open-source sketches in OpenProcessing
Thanks to OpenAI's Codex, I completed the migration of every single sketch into a clean, searchable, and structured GitHub repository in less than 10 minutes. Check the GitHub Site here.
Yes, really.
Image: GitHub folders after migration
Codex + GitHub = Seamless Integration
Codex, the AI model behind GitHub Copilot and part of OpenAI's ecosystem, isn't just for suggesting code. It can understand, refactor, and automate workflows with astonishing speed and accuracy. Here's how I did it:
Auth with GitHub: I connected my GitHub account via a secure OAuth integration. Within seconds, Codex had access to my account and repositories.
Scraping Sketches: Using Codex, I wrote a script that accessed each of my public sketches on OpenProcessing. It parsed the HTML and JavaScript content, organised filenames, and bundled them logically into folders—more than 500 folders and 30.000 lines of code in HTML and JavaScript.
Repo Structuring: Codex then helped me automate the creation of commit messages, generate readme files, and upload the organised sketches into versioned folders on GitHub.
Push and Done: With a final push to my GitHub repo, over 200 sketches were archived, portable, and open for collaboration or cloning.
GitHub Page: A page was created in less than 10 seconds.
Image: GitHub page with more than 200 clickable open-source sketches
Why This Matters
This isn’t just about saving time (though it did save hours, if not days). This is about empowering digital artists and creative coders to treat their work with the same level of organisation and permanence that software developers enjoy.
Portability: I’m no longer locked into a single platform. My sketches are now Git-tracked, forkable, and deployable anywhere.
Preservation: Online platforms can disappear or change. GitHub gives me version control and longevity.
Community Collaboration: With everything on GitHub, others can easily fork, remix, and build upon my work.
The Future
This experience is a perfect example of OpenAI's broader vision: integrating Codex and other models with third-party platforms in a way that makes advanced automation available to everyone, not just developers.
Whether you're a creative technologist, an educator, or a hobbyist, the ability to move, manage, and scale your projects with a simple natural language prompt is a total game-changer.
With just a handful of commands and some minimal oversight, I was able to:
Extract and process hundreds of files
Standardise naming conventions
Migrate them into Git-based workflows
Share my portfolio with the world in a developer-friendly way
This is a glimpse at what’s possible when AI becomes a true creative partner. The ability to bridge tools like OpenProcessing with platforms like GitHub using Codex is incredibly powerful and beautifully simple. This is the kind of technology that redefines how we work, create, and share.
And the best part? It took less time than a coffee break.