How to search your X (Twitter) bookmarks

What X gives you (and where it stops)

The bookmarks tab has a search box, but it's shallow — it matches words that literally appear in the post, and nothing else. Saved a thread about churn that never uses the word “churn”? Invisible. Remember the chart but not the author? Scroll and pray. There's no filtering by topic, date range, or author, and Premium's folders only help if you filed things when you saved them — which past-you didn't.

Workaround 1 — search operators on your own likes/saves

X's main search supports operators (from:, min_faves:, quoted phrases), but none of them scope to your bookmarks. The common trick — re-finding a post via global search and hoping you recognize it — is really just searching the whole firehose. It works when you remember exact words; it fails for the “that idea about pricing, somewhere” case, which is most cases.

Workaround 2 — build a searchable archive

Export your bookmarks (see every export method) and put them in Notion, Obsidian, or a spreadsheet. Now you have real full-text search — over a snapshot. The archive decays the moment you keep bookmarking, and keyword search still can't answer conceptual questions: it finds words, not meaning.

The actual fix — semantic search over a live sync

The question you want to ask is “what did I save about cold outreach?” — a meaning question, not a keyword question. That requires two things: a sync that keeps your bookmark set current, and AI that has read everything in it.

IdeaCoach does both: the Chrome extension keeps your bookmarks synced, every save gets an AI summary and category, and the chatlets you interrogate the whole pile in plain English — “what did I save about audience building?”, “which saved threads mention pricing experiments?”. It answers from your actual saves, with the source bookmarks attached.

Pick by the question you ask most

• “I remember the exact phrase”— X's native bookmark search is fine.
• “I need it once, for the record” — a one-time export + your notes app.
• “What do I know about X?” — semantic search over a live sync is the only thing that answers it.