LLM Knowledge Bases: a practical guide — Ben Holmes, Warp

A personal knowledge base can start with rough dictated notes and acquire structure afterward. The workflow description separates capture from enrichment: an agent adds tags from a fixed vocabulary, researches sources and links related notes.

Timestamps mark files already processed, while a generated wiki provides browsable entries for concepts and other subjects mentioned in the notes. Scheduled processing synchronizes files into a cloud environment, enriches new material and sends the updates back. The organizing work therefore becomes a repeatable pass over captured material, reducing the need to classify every thought while recording it.

“Talking runs about 200 words a minute and typing does not, which is why Ben Holmes starts a knowledge base with voice dictation rather than with organization.”

AI Engineer

“The pipeline only works if there is enough raw material, so the capture step is deliberately sloppy: ramble into your phone after a podcast or a meeting, skip the formatting, and let agents impose structure afterward.”

AI Engineer

“An enrich note skill stamps each file with a timestamp so later runs know what has been processed, assigns tags from a fixed reference list instead of inventing new ones, since as he puts it Claude loves to get creative, researches the source on the web, and adds backlinks found by key term search.”

AI Engineer

“The closing trick is a graph view of every note, built in HTML and Tailwind purely by asking an agent for it, which turns out to be a decent map of where your own thinking has gaps.”

AI Engineer