A Claude-powered folder turns raw files into a self-updating second brain

wschroll · x · 2026-07-29

A long post argues that the real value of a Claude-powered “second brain” is not the visual graph, but the questions it can answer from your own files.

The workflow is simple: drop raw documents into one folder, let Claude read them, split them into notes, connect related ideas, and keep the graph updated automatically. The author says this lets you ask things like:

Because the answers are grounded in your own writing with dates and sources, it avoids the usual problem of chatting with a model that forgets you every 24 hours.

Original post →

More from coding & agent

coding & agent channel →