Cairn

Cairn is a small personal knowledge base for my X bookmarks and likes.

I had started using X as a read-it-later placeholder. The saved posts, threads and articles carried a lot of useful signal, but a few thousand saves get hard to work with once they pile up. Cairn pulls them into a database so they are easier to search, organize, chat with and revisit.

How it works

Cairn imports your saved posts and likes into a local database, then layers search and organization on top. The import side is the harder half of the problem, and I wrote it up separately in Retrieving likes and bookmarks from x.com: the auth, pagination, archive parsing and rehydration involved in getting your own saves out of X.

The knowledge graph

Behind the scenes, Cairn connects saved posts, links, authors, topics, repos, articles and papers into a lightweight knowledge graph. One save can lead to related posts and forgotten adjacent ideas, so the collection reads less like a flat list and more like a map of what you have been paying attention to.

Chat with your saves

Cairn supports LLM chat with bring-your-own-key. You point it at your own model and key, then ask questions across everything you saved, without handing your reading history to a service you do not control.

Cairn is at an early, alpha stage and built for my own use. A short walkthrough is in the demo video.