My current workflow to create the weekly reading posts 🔗
Here is a breakdown of the simple, automated, and edited workflow I currently follow every week:
1. Capturing and Collecting Links 🔗
I use Raindrop.io as a Google Chrome extension and mobile app to save the links I found interesting into a dedicated collection named “00-current”. This collection is configured to be the default saving location for all new bookmarks.
2. Processing (Sunday) 🔗
- I export all the bookmarks saved in the “00-current” collection as a .txt file.
- I use Gemini to create the initial structured Markdown post. I input the exported links and instruct the model to:
- Organize the links into relevant technical categories (Kubernetes, Cloud Native, DevOps, etc.).
- Generate a brief, descriptive summary for each link.
- Format the entire output using the standard Markdown required by this blog.
3. Review, Refine, and Publish 🔗
- I review and edit the AI-generated markdown draft to prevent hallucinations and ask Gemini to refine the post if so.
- I use the GitHub web editor (yes, I’m lazy) to create a PR with the final markdown file.
- I review the preview (I currently use Netlify for that) and refine it if needed.
- I merge the PR and the post is published (via GitHub actions)
To prepare for the next week, I move all the links that were just published from “00-current” to a collection named “99-archive” in raindrop for archiving purposes.
That’s it!
