How search works (and what it deliberately can’t find)

Search looks across entry names, entry fields, and the text inside your documents — and it all happens in your browser, after your vault is unlocked. Your searches never leave your device.

Deliberately excluded

Some fields never enter the search index at all: passwords, Social Security numbers, account and routing numbers, PINs, card numbers, seed phrases, safe combinations, dollar balances, and other high-sensitivity fields. That’s a design decision, not a bug — a search index is a copy, and those values are too sensitive to copy anywhere, even inside your own encrypted vault session. So searching for the last four digits of an account number finds nothing; search for the entry’s name (“Chase checking”) instead.

Document text

When you upload a real PDF (one with actual text in it), the text is extracted, encrypted with the same key as everything else, and becomes searchable. Two limits to know:

• Scanned photos have no text layer — a photographed page can’t be text-searched. The app tells you when a document has no extractable text. See Scanning documents.
• Very long documents are indexed up to roughly the first 50 dense pages; text beyond that isn’t searchable.

In both cases, the fix is the same: give documents descriptive names. Names are always searchable.