Import by identifier

The problem

You have a list of DOIs (or ISBNs, or arXiv IDs) from a reading list, a syllabus, or a paper’s references, and you want them in Zotero with full metadata and PDFs. Clicking each one through the browser connector is slow, and it doesn’t script.

Via an AI agent

“Add these three DOIs to my To Read collection with their open-access PDFs: 10.1371/journal.pmed.0020124, 10.1145/3442188.3445922, 10.1038/s41586-021-03819-2.”

The agent runs one zot add per identifier, attaches the OA PDF when available, and files each into the collection.

Via the CLI

zot add doi 10.1371/journal.pmed.0020124 --pdf                 # DOI + open-access PDF
zot add isbn 9780262035613 --collection "To Read"              # ISBN into a collection
zot add arxiv 2103.00020 --pdf --collection "To Read"          # arXiv + PDF

--pdf fetches an open-access PDF when one is available; --collection files the new item directly. Metadata is resolved from the identifier, so the imported item lands with title, authors, date, and DOI already populated.

Scripting a whole list:

while read -r doi; do
  zot add doi "$doi" --pdf --collection "To Read" --yes
done < dois.txt

Safety net

Imports are additive — they create new items and never modify or delete existing ones, so they are low-risk. If an import is wrong, send the new item to the trash (recoverable) rather than erasing it. Run zot missing doi or zot lint afterward to confirm the new items came in clean, and see Find missing metadata to fill any gaps.