Zotero MCP
Used as a local embedding backend for semantic search over the Zotero library.
Used for semantic search over the library via OpenAI embedding models, complementing local sentence-transformer embeddings.
Connects to a user's Zotero research library to search it (by title, author, tag, collection, full text, or semantic meaning), read item metadata, BibTeX, full text and PDF page ranges/images, create and extract highlights, area annotations and notes, and write back to the library (adding papers by DOI, URL, ISBN, BibTeX or file, managing collections and tags, merging duplicates). Works against the local Zotero installation/zotero.sqlite in local mode or the Zotero web API, and ships a companion CLI plus a Zotero plugin that adds an in-app chat panel aware of the open item, page and selection.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Zotero MCP find my papers on diffusion models and summarize the key findings"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Zotero MCP connects your Zotero research library with ChatGPT, Claude, and other AI assistants (e.g., Cherry Studio, Chorus, Cursor) via the Model Context Protocol. Search your library, read and annotate papers, add and organize items, and find research by meaning.
AI agents: read docs/for-agents.md first. It covers which route to use, setup, and the commands in one place.
Three ways in
One package, zotero-mcp-server, ships all three. Pick the one that fits where you work.
MCP server |
| Zotero Agent | |
For | Chat apps that speak MCP but have no shell: Claude Desktop, ChatGPT, Cherry Studio, Chorus | Agents with a shell: Claude Code, Codex, Cursor, Windsurf, Gemini CLI, Amp, OpenCode | Chatting inside Zotero, next to the PDF you are reading |
How |
|
|
|
Why | Works in any MCP client | 98 tokens in context until it is needed, instead of ~8k | Knows the item, page and selection you have open; answers link to pages |
Guide |
Related MCP server: Zotero MCP
What it does
Search by title, author, tag, collection, full text, or meaning (semantic search with local, OpenAI, Gemini, or Ollama embeddings)
Read metadata, BibTeX, full text, and page ranges of PDFs, with page images where text extraction garbles math, figures, and tables
Annotate: highlights and area boxes placed on the exact words, figure, table, or equation; notes; PDF annotation extraction
Write: add papers by DOI, URL, ISBN, BibTeX, or file (with open-access PDFs), manage collections and tags, merge duplicates
Local or web: in local mode reads come straight from
zotero.sqlite; writes go to the running Zotero 10+ or through the web APIScite citation tallies and retraction alerts (optional)
Quick start
1. Install (Python 3.10+):
uv tool install zotero-mcp-server # or: pip install zotero-mcp-serverNew to the command line? Try the community-built Zotero MCP Setup: a macOS GUI installer, one-click scripts for Mac and Windows, and a step-by-step guide.
2. Enable Zotero's local API: in Zotero 7+, open Settings → Advanced and tick Allow other applications on this computer to communicate with Zotero.
3. Connect your assistant:
zotero-mcp setup # auto-configures Claude Desktopor add the server by hand to claude_desktop_config.json:
{
"mcpServers": {
"zotero": {
"command": "zotero-mcp",
"env": { "ZOTERO_LOCAL": "true" }
}
}
}Teach the agent to drive zotero-cli (the cheaper route, see below):
zotero-mcp install-skillOr use the MCP server: for Claude Code, add the same mcpServers entry as above to ~/.claude.json.
See Getting started.
zotero-cli plugin # where the .xpi is, and how to install itThen in Zotero: Tools > Plugins, the gear, Install Plugin From File. Details: Zotero Agent plugin.
4. Writes (optional): on Zotero 10+, run zotero-mcp authorize-local once and choose Always Allow. On older Zotero, add ZOTERO_API_KEY and ZOTERO_LIBRARY_ID to write through the web API.
Then ask things like "Find papers in my library on attention mechanisms", "Summarize the key findings of this paper", or "Highlight the main claims of this PDF".
Optional extras
The base install covers search, reading, annotations, and writes. Heavier features are extras:
Extra | What it adds | Install command |
| Semantic search via ChromaDB, sentence-transformers, OpenAI/Gemini embeddings |
|
| PDF outlines, page layout and page images (PyMuPDF), EPUB annotations |
|
| Scite citation tallies and retraction alerts (no account needed) |
|
| Everything above |
|
Update any time with zotero-mcp update.
MCP server or agent skill?
If your agent has a shell (Claude Code, Cursor, Codex, Windsurf, Gemini CLI, Amp, OpenCode …), one command teaches it to drive zotero-cli:
zotero-mcp install-skillAn MCP server sends every tool's schema on every request, before you type anything. The skill costs 98 tokens until the agent decides it is relevant:
Route | In context | Paid |
MCP server, default profile (38 tools) | 7,939 | every request |
MCP server, all tools (50) | 11,225 | every request |
Agent skill, frontmatter only | 98 | always |
Agent skill, body loaded | 2,400 | when it fires |
Agent skill + full command reference | 6,582 | worst case |
Use the MCP server when your client speaks MCP but has no shell (Claude Desktop, ChatGPT); use the skill when it has a shell. Both share one config. Details: CLI and agent skill.
Zotero Agent: chat inside Zotero
A plugin that adds a chat panel to Zotero itself, driven by your own Claude Code, Codex, pi or OpenCode. It knows which item and page you have open and works your library through zotero-cli. Answers cite papers with real zotero:// links, so a click opens the PDF at that page. It ships in the wheel: run zotero-cli plugin for the file and the install steps, or paste the one-paragraph prompt in the docs to your agent and let it do the setup. Details: Zotero Agent plugin.
Documentation
Guide | What's in it |
Connecting Claude Desktop and Claude Code, VS Code / GitHub Copilot, ChatGPT, Cherry Studio, Chorus, Autohand, and other MCP clients | |
Environment variables, local writes, web and hybrid modes, the SQLite read backend, global search, text extraction, command-line options | |
Embedding models, building and updating the index | |
Every MCP tool, tool groups ( | |
| |
The chat panel inside Zotero: install, first run, agents and sign-in, privacy | |
Container images and runtime modes | |
Common problems and fixes | |
One guide for an agent setting up or using Zotero MCP |
Website: stevenyuyy.com/zotero-mcp · Changelog
Contributing
Issues and pull requests are welcome. Run the tests with uv run pytest tests/. A live integration test plan, meant to be run by Claude against a real library, is in docs/integration-test-plan.md. Changes to the plugin in plugin/ (AGPL-3.0) are accepted on the condition that the maintainer may also release them under other terms, including commercial ones.
Thanks to everyone who has contributed code, fixes, and ideas to Zotero MCP.
Support
Zotero MCP is free and open source.
If it saves you or your lab time, sponsoring helps cover the unglamorous parts: Windows and WSL2 edge cases, Zotero schema changes, group-library support, and the embedding/search infrastructure.
Labs and institutions: the $50 and $200 tiers are meant to be expensable, and include priority triage on the issues affecting your workflow.
License
The MCP server and zotero-cli are MIT (LICENSE). The Zotero Agent plugin in plugin/ is AGPL-3.0-or-later (plugin/LICENSE), the license Zotero itself uses.
Available Tools
35 toolszotero_add_itemZotero Add ItemA
Add items from a DOI, URL, ISBN, PMID, BibTeX, CSL JSON or a local file; use it for every 'add this to Zotero' request. source: the identifier, URL, citation text or ABSOLUTE file path. DOI/URL/ISBN/PMID also take a comma- or newline-separated batch; BibTeX and CSL JSON may be inline (many entries) or a .bib/.json path. source_type: 'auto' (default) detects it. Prefer a DOI: a URL other than doi.org or arxiv.org becomes a bare webpage item, and ISBN metadata is noisy (verify it). For a file, its DOI is looked up, else metadata is guessed, then the file is attached; title overrides a missed title. collections: keys, names or 'Parent/Child' paths, checked before anything is created (create_missing_collections=True makes missing ones). if_exists: 'file' (default) reuses an item with the same DOI/arXiv ID/ISBN/URL, adding missing collections and tags; 'duplicate' always creates; 'skip' leaves a match alone. attach_mode: 'auto' (default) attaches an open-access PDF, 'linked_url' links it, 'none' skips, 'required' fails without one.
Inputs: source (required), tags, title, if_exists, +4 more.
zotero_advanced_searchZotero Advanced SearchA
Search with structured field conditions, for what zotero_search_items cannot express: date ranges, item types, matches on one field. conditions: a list of {field, operation, value}. Fields include title, creator, date, dateAdded, dateModified, tag, itemType, publicationTitle, abstractNote, collection. Operations: is, isNot, contains, doesNotContain, beginsWith, endsWith, isGreaterThan, isLessThan, isBefore, isAfter. Dates are ISO; 'added in the last 30 days' is dateAdded isAfter . join_mode: 'all' (AND, default) or 'any' (OR). sort_by: a field such as dateAdded or title; sort_direction: 'asc' (default) or 'desc'. limit: default 50, max 500. include_subcollections=True makes collection is/isNot match the whole subtree. search_all_libraries=True searches every library (no collection conditions). Example: conditions=[{'field': 'itemType', 'operation': 'is', 'value': 'preprint'}, {'field': 'dateAdded', 'operation': 'isAfter', 'value': '2026-03-22'}].
Inputs: conditions (required), limit, sort_by, join_mode, +3 more.
zotero_attach_fileZotero Attach FileA
Attach a file to an EXISTING regular item (not an attachment or note). To create a new item from a file, use zotero_add_item(source=). Give exactly one of file_path (absolute; any regular file, no symlinks or .lnk) or url (a direct PDF link, downloaded by the server). filename: overrides the stored name. link_mode: 'import' (default) uploads a copy, which counts against Zotero storage unless WebDAV is set up; 'linked' links the file where it is (file_path only, nothing copied). Skips the upload when the item already has the same file. Returns the attachment key.
Inputs: item_key (required), url, filename, file_path, +1 more.
zotero_authorize_local_writesZotero Authorize Local WritesA
Ask the user to allow writes to the local Zotero library (Zotero 10+). BLOCKS until they answer a dialog in the Zotero app, so tell them to switch to Zotero first. 'Always Allow' saves a reusable key; 'Allow' covers one write. Call it when a write fails and zotero_write_capabilities says local writes are supported but no key is held. app_name: shown in the dialog. timeout: seconds to wait (5-55, default 45). Zotero allows about 5 prompts a minute: do not retry in a loop.
Inputs: timeout, app_name.
zotero_batch_updateZotero Batch UpdateA
Edit many items in one call: add or remove tags, and set or remove Key: value lines in Extra (BetterBibTeX keys, tex.* fields). Select by item_keys, or by query and/or tag (ANDed; a list of tags ORs); item_keys wins. Needs a selector and an action. add_tags/remove_tags leave other tags alone. set_keys upserts Extra lines by case-insensitive key; remove_keys deletes them. limit: items selected by query/tag (default 50, up to 1000); the report says if more matched. Example: (tag='to-read', add_tags=['reviewed'], remove_tags=['to-read']).
Inputs: tag, limit, query, add_tags, +4 more.
zotero_create_annotationZotero Create AnnotationA
Annotate a PDF (EPUB: highlights and underlines only). One mode per call: text= highlights that exact text (underline=True underlines it), matched against the text layer (scanned PDFs will not match); rect=[x, y, width, height] draws an area box over a figure or table; note=[x, y] places a sticky note at that point, its text in comment. rect and note are PDF only, normalized 0-1 from the page's top-left: take boxes from zotero_get_page_layout instead of guessing. attachment_key: the attachment, NOT the parent item (zotero_get_item_children lists it). page: 1-indexed (EPUB: chapter). comment, color (hex, default '#ffd400') and tags are optional. Needs PyMuPDF (the [pdf] extra). Example: (attachment_key='NHZFE5A7', page=7, rect=[0.15, 0.22, 0.6, 0.35], comment='Figure 3').
Inputs: page (required), attachment_key (required), note, rect, +5 more.
zotero_delete_annotationZotero Delete AnnotationA
Permanently delete a Zotero annotation. This cannot be undone. Annotations are deleted outright, as Zotero's own PDF reader does: a trashed annotation stays visible in the reader with no way to remove it there.
Inputs: annotation_key (required).
zotero_delete_itemZotero Delete ItemA
Move an item to the Trash (recoverable in Zotero). Any item type, but for notes use zotero_manage_note(action='delete'), or pass allow_note=True.
Inputs: item_key (required), allow_note.
zotero_export_bibliographyZotero Export BibliographyA
Format references with Zotero's CSL engine. item_keys: up to 100 item keys (takes precedence); or collection_key; neither exports the active library's top-level items. style: a CSL style short name (default 'apa'; e.g. 'modern-language-association', 'chicago-note-bibliography', 'ieee'). export_format: 'bib' (reference list, default), 'citation' (in-text citations) or 'bibtex'.
Inputs: style, item_keys, export_format, collection_key.
zotero_get_annotationsZotero Get AnnotationsA
Get PDF/EPUB annotations (highlights and their comments) for one item, or for the whole active library when item_key is omitted (can be thousands, so pass item_key when you know it). item_key: the parent item or an attachment. use_pdf_extraction=True parses the PDF itself when Zotero has no stored annotations (e.g. ones made in another reader). limit: default all. format: 'markdown' (default) or 'json' (normalized records).
Inputs: limit, format, item_key, use_pdf_extraction.
zotero_get_attachment_pathZotero Get Attachment PathA
Return the local filesystem path(s) of a Zotero item's attachments. Local mode only. Useful when you want to read a large PDF directly (e.g., a book) instead of going through zotero_get_item_fulltext, which is page-limited.
Inputs: item_key (required).
zotero_get_collection_itemsZotero Get Collection ItemsA
List the items in one collection. detail: 'keys_only', 'summary' (default, no abstracts) or 'full' (with abstracts). include_subcollections=True adds items in nested collections. Page with limit (default 50) and offset; the response names the next offset. To find papers on a topic, zotero_semantic_search is better.
Inputs: collection_key (required), limit, detail, offset, +1 more.
zotero_get_collectionsZotero Get CollectionsA
List every collection in the active library as an indented tree with keys. To find one collection by name, zotero_search_collections is cheaper. limit: default 100, up to 5000. include_trashed=True also shows collections in the Trash.
Inputs: limit, include_trashed.
zotero_get_item_childrenZotero Get Item ChildrenA
List the child attachments and notes of one or more items. Use it to get the attachment key that zotero_create_annotation and zotero_get_page_layout need. item_key: one key or an array of keys; pass all of them in one call (a bad key is reported, not fatal). One key shows attachments (type, filename) and notes in full; several show one line per child, grouped by parent.
Inputs: item_key (required).
zotero_get_item_fulltextZotero Get Item FulltextA
Return the extracted text of an item's PDF or EPUB: most of the paper, often 10K+ tokens. Use it only when the user wants the paper READ, never to search (zotero_semantic_search, zotero_search_items) or for metadata (zotero_get_item_metadata); avoid reading several papers unless asked. item_key: normally the parent item, whose best attachment is picked (PDF first unless attachment_priority says otherwise); an attachment's own key reads exactly that file. PDFs stop at fulltext_display_max_pages (default 10), and a TRUNCATED heading names the pages read: continue with zotero_read_pdf_pages. Scanned PDFs without OCR may return little text.
Inputs: item_key (required).
zotero_get_item_metadataZotero Get Item MetadataA
Metadata for ONE item by its 8-character key (not a DOI or title): title, creators, date, DOI, tags, abstract, URL, collections (as keys). format: 'markdown' (default), 'json' (the raw Zotero record) or 'bibtex'. include_abstract=False shortens markdown output. Trashed items are returned too, marked In Trash. To read the paper use zotero_get_item_fulltext; to find items, the search tools.
Inputs: item_key (required), format, include_abstract.
zotero_get_notesZotero Get NotesA
Read notes in the active library. Without query it LISTS notes: an item's child notes with item_key, else notes library-wide. With query it SEARCHES note and annotation text (case-insensitive substring, library-wide, so not with item_key). limit: default 20. truncate=False returns whole bodies (list mode). raw_html=True returns the note's HTML, for editing and passing back to zotero_manage_note(action='update').
Inputs: limit, query, item_key, raw_html, +1 more.
zotero_get_page_layoutZotero Get Page LayoutA
Detect figure, table and display-equation regions on one PDF page and return their boxes (x, y, width, height, normalized 0-1), each with its caption or equation number, a confidence and a ready-made zotero_create_annotation call. Use it before an area annotation instead of guessing coordinates. Boxes cover the graphic itself: labels inside a figure or unruled table headers can fall outside, and confidence reflects caption matching, not box completeness. attachment_key: the PDF attachment, NOT the parent item (zotero_get_item_children lists it). page: 1-indexed. PDFs only.
Inputs: page (required), attachment_key (required).
zotero_get_pdf_outlineZotero Get Pdf OutlineA
Return a PDF's table of contents with page numbers, usually under 200 tokens: find the pages you need, then read them with zotero_read_pdf_pages instead of the full text. item_key: the parent item or the PDF attachment. Says so when the PDF has no outline. PDFs only; needs PyMuPDF (the [pdf] extra).
Inputs: item_key (required).
zotero_get_recentZotero Get RecentA
List the most recently ADDED items, newest first, in the active library or one collection (collection_key). For 'what did I add lately', not topic search or a collection's full contents. Includes standalone notes and attachments. limit: default 10.
Inputs: limit, collection_key.
zotero_get_tagsZotero Get TagsA
List every tag in the active library, one per line. Use it to find tag names before filtering with zotero_search_items(tag=...) or editing with zotero_batch_update. limit: default all.
Inputs: limit.
zotero_list_librariesZotero List LibrariesA
List the libraries you can switch to: the personal library, group libraries (with group IDs) and, in local mode, RSS feeds, each with its item count. Use it to get an ID for zotero_switch_library. The active library is not marked.
Inputs: no named parameters.
zotero_manage_collectionZotero Manage CollectionA
Create, update or delete a collection. action='create': name, and optionally parent_collection for a subcollection. 'update': collection_key plus a new name, a new parent_collection, or to_top_level=True; the key, items and subcollections are kept, and a collection cannot move under itself or its own subcollection. 'delete': collection_key; permanent (Zotero does not trash collections) and takes its subcollections with it, but the items stay in the library. parent_collection: a key, name or 'Parent/Child' path. To file items into collections use zotero_set_item_collections.
Inputs: action (required), name, to_top_level, collection_key, +1 more.
zotero_manage_noteZotero Manage NoteA
Create, update or trash a note. item_key: the PARENT item for action='create'; the NOTE's own key for 'update' and 'delete' (zotero_get_notes finds it). note_text: Markdown with $math$ plus (or background-color), or note HTML (sanitized). note_title becomes a heading and tags are optional (create). update REPLACES the body unless append=True; to keep formatting, edit the HTML from zotero_get_notes(raw_html=True) and pass it back whole. delete moves the note to the Trash (recoverable). Notes only: other items go through zotero_delete_item.
Inputs: action (required), item_key (required), tags, append, +2 more.
zotero_read_pdf_pagesZotero Read Pdf PagesA
Read chosen pages of an item's PDF, e.g. after zotero_get_pdf_outline. item_key: the paper or its PDF attachment. Reads start_page through end_page (1-indexed; omit end_page for one page). format='text' (default) returns Markdown and flags pages whose equations, figures or tables the text garbles; format='image' returns PNG page images (up to 10) to read those exactly. rect=[x, y, width, height] (normalized 0-1, e.g. from zotero_get_page_layout) returns just that region of start_page, zoomed in.
Inputs: item_key (required), start_page (required), rect, format, +1 more.
zotero_search_by_citation_keyZotero Search By Citation KeyA
Find the one item with this BetterBibTeX citation key (exact, case-sensitive), e.g. 'smith2024'. Checks Zotero's citationKey field, then 'Citation Key:' lines in Extra. For partial keys or free text use zotero_search_items.
Inputs: citekey (required).
zotero_search_collectionsZotero Search CollectionsA
Find collections by name in the active library; returns their keys and parents. Case-insensitive substring on the collection's own name, and every word must match ('reading list' needs both). include_trashed=True also matches collections in the Trash.
Inputs: query (required), include_trashed.
zotero_search_itemsZotero Search ItemsA
Keyword search of item metadata: substring match on title, creators and year (qmode='everything' adds abstracts). Keep queries short, like 'Brewer 2011' or a surname: every extra word narrows the match. For topics or concepts use zotero_semantic_search. If nothing matches it retries simpler queries, then semantic search, and says so. tag: conditions that are ANDed; within one, ' OR ' alternates and a leading '-' excludes, e.g. ['methods OR methodology', '-draft']. Omit query to list items by tag alone (zotero_get_tags lists the names). item_type: '-attachment' (default) excludes attachments, or e.g. 'journalArticle'. limit: default 10. collection_key: search one collection (no retries); include_subcollections adds its subcollections. search_all_libraries=True searches every library and labels each result; needs a query, not with collection_key.
Inputs: tag, limit, qmode, query, +4 more.
zotero_set_item_collectionsZotero Set Item CollectionsA
Add items to, or remove them from, collections, leaving their other memberships alone. item_keys: an array of item keys. add_to / remove_from: arrays of collection keys, names or 'Parent/Child' paths, all checked before anything changes. To replace memberships wholesale use zotero_update_item; to create, rename or delete a collection, zotero_manage_collection.
Inputs: item_keys (required), add_to, remove_from.
zotero_set_item_parentZotero Set Item ParentB
Set or clear the parent of a Zotero item. Pass a parent item key to assign or change the parent, or null to make the item top-level. Zotero validates whether the requested parent-child relationship is allowed.
Inputs: item_key (required), parent_key (required).
zotero_switch_libraryZotero Switch LibraryA
Set the library that every later tool call reads and writes, for the rest of the session. library_id: an ID from zotero_list_libraries ('user' or '0' for the personal library). library_type: 'group' (default), 'user', 'feeds', or 'default' to go back to the configured library (library_id is then ignored). Fails if the library is not accessible.
Inputs: library_id (required), library_type.
zotero_synthesize_annotationsZotero Synthesize AnnotationsA
Gather highlights, comments and child notes across a scope into a per-paper digest for YOU to synthesize; no LLM is called. collection_key: only papers in that collection; omit for the whole active library. tag: a tag or list of tags to filter papers by. limit: annotations and notes scanned (default 200). format: 'markdown' (default, grouped by paper with a count summary) or 'json'. Use it before writing a thematic review.
Inputs: tag, limit, format, collection_key.
zotero_update_annotationZotero Update AnnotationA
Update an existing Zotero annotation. Editable fields: text (highlight text), comment, color (hex like '#ffd400'), and tags. Tags can be replaced wholesale via tags, or edited incrementally via add_tags/remove_tags (mutually exclusive with tags). Position/page/sortIndex are anchored to the PDF/EPUB geometry and are not editable.
Inputs: annotation_key (required), tags, text, color, +3 more.
zotero_update_itemZotero Update ItemA
Change metadata on one item; only what you pass changes. fields: {name: value} with snake_case names (title, date, doi, url, abstract, publication_title, short_title, book_title, citation_key, item_type, place, extra, volume, issue, pages, publisher, issn, isbn, edition, language, access_date) or raw Zotero field names. Unknown names fail and list the valid ones; names the item's type lacks are skipped and reported. citation_key sets the BetterBibTeX key; item_type converts the item, dropping fields the new type lacks. creators: a full replacement list of {creatorType, firstName, lastName}. tags REPLACES all tags; add_tags/remove_tags edit incrementally (preferred; not with tags). collections (keys) or collection_names REPLACE membership (collections=[] clears it); for incremental moves use zotero_set_item_collections. For notes use zotero_manage_note.
Inputs: item_key (required), tags, fields, add_tags, +4 more.
zotero_write_capabilitiesZotero Write CapabilitiesA
Report whether this server can write to Zotero (local API, web API, hybrid, or not at all) and what is missing. Instant; opens no dialog. Call it when a write tool reports no writable backend, before asking the user: if local writes are supported but no key is held, call zotero_authorize_local_writes; otherwise web API credentials are needed.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 35 tools
The tool set is broad, but descriptions generally route clearly between close areas: notes vs annotations vs synthesized digests, keyword vs advanced vs citation-key search, and item vs collection membership edits. A few boundaries remain fussy, especially update_item vs batch_update vs set_item_collections, and references to a missing semantic search add some confusion.
Almost all tools use the zotero_ prefix, snake_case, and a predictable verb-first pattern such as get_, search_, create_, update_, delete_, manage_, and list_. Minor deviations like zotero_write_capabilities and multi-action manage_ names are still readable and consistent enough.
35 tools is heavy for one MCP server, even for a broad domain like Zotero. Several specialized operations such as set_item_parent, get_page_layout, get_attachment_path, and write_capabilities push the surface well past a well-scoped range, increasing selection overhead.
The surface covers item CRUD, collections, notes, annotations, PDF reading/outlining, bibliography export, library switching, and write authorization. The main gap is that multiple descriptions point agents to zotero_semantic_search, but that tool is not present in the provided set.
Maintenance
Related MCP Connectors
Personal context for every AI: search, read, and write back to your private Markdown library.
AI research library. Save, organise and reuse notes and webpages as clean markdown context.
Academic literature search, retrieval, and private library management on top of OpenAlex.
Research paper search with real citations and reference formatting for AI assistants
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceConnects your Zotero research library with AI assistants (Claude, ChatGPT, etc.) via the Model Context Protocol, enabling paper review, summaries, annotations, and semantic search both locally and through the web API.MIT
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to search a personal research library, retrieve complete item metadata, and read the full text of attached PDFs through the Model Context Protocol. It can connect either to the local desktop application's API or to the web API using an API key and library ID.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to search a Zotero library, retrieve complete item metadata, and access full text from stored PDFs through the Model Context Protocol.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to search a Zotero library, retrieve complete item metadata, and fetch full text such as PDF contents through the Model Context Protocol.-