February 14, 2026
Daily List Collections
As of today, there are 3466 articles posted in the daily list archive on my website.
That’s a lot to read through.
And there are at least a half dozen topics that I write about on a regular basis.
And at least a half dozen more sub-topics under each of those.
AND I don’t write in a serial format.
Instead, I choose what to write about based on whatever inspired me that day.
(Which is another way to say that I jump around)
So...
One of the things I’ve always wanted on my website was a way for visitors to create a curated list of just the articles they need to solve the main challenge they’re facing.
Tagging was my first thought, but a tag for a core theme like “pricing” or “positioning” would still return thousands of posts.
That’s too many.
So I turned to Codex to see what it would come up with.
I started a new project, pointed it at the root directory of my local website repo, and said:
This project contains the source code and data for my website. There’s a folder in it named
dailythat has about 3500 markdown files in it. They are articles I sent to my mailing list, which I also archive as blog posts. Currently, they are listed on my site only in chronological order. I would like to develop a way for website visitors to find collections of articles that are useful for whatever situation they’re facing. For example, a user might say, "How do I get more leads for my dev shop?" and my website would present a collection of articles that would help. I don’t want you to edit or move anything in the project yet. Just give me your thoughts on how it might work.
I went on to give it constraints like:
- Do not change anything in or about the existing files
- Build this as a free-standing app that lives at the top level of the repo
- You can’t use a database; static files only
- The articles must be indexed by what problem they solve, NOT by what strings they contain
About an hour later, I had this:
It’s not perfect, and I’m going to keep iterating on it, but it’s WAY better than the existing search.
If you give it a try, plmk how I could make it better for your use case.
Yours,
—J