Zotero attachments, full text, sync and export

zotero_groups

Lists the group libraries your key can access (id, name, type, item count, edit permissions). Pass a returned id as library_id (with library_type:"group") to other tools to operate on a group library.

zotero_export

Exports items in a machine-readable bibliographic format and returns the raw text: bibtex, biblatex, ris, csljson, csv, mods, tei, coins, rdf_bibliontology, rdf_dc, rdf_zotero, refer, wikipedia, bookmarks. Narrow with item_keys, collection_key, q, or item_type. A limit is always applied (export formats require it). For styled, human-readable citations in a CSL style (APA, IEEE, …), use the citation tools.

zotero_fulltext

Read or write attachment full text (only attachment items have it):

• get — extracted text + indexing stats (found:false when none).
• set — store extracted text (content + char/page counts).
• since — map of attachment keys whose full text changed after a library version (for incremental indexing).

zotero_sync

The version-based delta the Zotero sync algorithm uses. Given since (a library version, 0 = everything), returns per-type maps of changed keys (items/collections/searches/tags) plus the deletion log. Fetch only the changed keys afterward — don't re-pull the whole library.

zotero_attachment

Upload, download, or inspect attachment files. File bytes go to/from disk, never through the conversation.

• upload — store a local file via the full 5-step Zotero File Storage protocol (compute md5/mtime → request authorization → upload bytes → register). Optional parent_item, title, content_type. Returns the new attachment key (and whether the file already existed in storage).
• download — fetch an attachment's file to save_path (default under the Zoteus data dir); returns the path and byte count.
• info — return an attachment item's metadata.

Uploads/downloads use the cloud Web API and count against your Zotero file-storage quota.