the pile
Pull up your X bookmarks. Scroll. Keep scrolling. If you've been on the app for more than a year, you have somewhere between a few hundred and a few thousand saved posts. You don't remember most of them. You've never gone back. You will probably never go back.
This is fine. Bookmarks are cheap. The cost of saving is one tap; the cost of revisiting is infinite. So the pile grows. The pile is also, accidentally, the most honest record you keep about yourself.
what the pile knows
The pile knows what caught your eye when you were tired. What you wanted to think about later but didn't. What you thought might be useful in a future you haven't reached. There's a version of you that lives in there — quieter than your timeline, less curated than your likes, more honest than your follows.
A person's bookmarks tell you:
- What they aspire to read (the books, the longform, the threads).
- What they're building (the threads on stack choice, the technique posts).
- What they wish they understood (the threads they saved twice, in different forms).
- Who they trust (the same handles, year after year, in the citation graph).
the problem with the pile
The pile is unreadable. It's sorted by save date, not relevance. It has no full-text search worth the name. It can't handle a question like “what did I save about agents in November?” or “the thread that explained CRDTs.” The native app is good at capture and bad at recall.
That asymmetry is what xmark fixes. We pull your bookmarks, embed them, and let you ask questions in plain English. The librarian doesn't care when you saved something. The librarian cares what you meant.
what changes when you can read the pile
People who use xmark for a few weeks tell us the same thing: their bookmarks stop being a graveyard and start being a library. They notice patterns they didn't know they had. They quote a tweet from eight months ago in a doc and find it in twelve seconds.
Your bookmarks are already a personality. xmark just gives you a way to read it back.