Best "Brain" for Agents Is Just Versioned Folders of Markdown Files

I think just one useful thing is mentioned: put your md files in git and then have a hybrid fts/vector search across them actually does work better? Not a very surprising conclusion. Doesn't need that much text to explain, does it?

I’m confused: is it just markdown files in git? Or does the hybrid graph+semantic layer matter? If the latter is true, the title is just clickbait right

This should be common sense and immediately obvious to anybody who has spent more than a few hours with a coding agent.

How does git replace a vector db search exactly? They are orthogonal. Are you gonna burn a million tokens every time you wanna find some relevant files?

This isn't even well-slopped slop.