A blog about building
your own LLM minibases.
LLMs are only as smart as the context you give them. A minibase is a small, curated knowledge base on a topic you actually care about, structured so that Claude, ChatGPT, or whatever model you use can read it and think with you.
Latest posts
-
What are minibases
A working definition, a worked example, and why one big knowledge base stops being enough once you have more than one topic you actually care about.
-
Obsidian vs. Save vs. plain folders: pick the one you'll actually maintain
A short, opinionated tour of the three tools I keep recommending, which one fits which kind of person, and how to migrate without losing your mind.
-
Web clipping for LLMs: Obsidian Web Clipper, Save, and a Reader pipe
The three capture flows that actually produce useful files. Not the PDF dumps, not the screenshots, not the bookmarks app. The ones that give the model clean prose to work with.
Stay in the loop
No newsletter yet, but there will be one soon. In the meantime the easiest way is to bookmark this page or follow along on X.