zotero-mcp
Provides read-only tools to search, retrieve metadata, export citations and BibTeX, list collections and tags, and access full text from a Zotero library via its local API.
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 papers on reinforcement learning in my Zotero"
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
A small, read-only MCP server that exposes your Zotero library to Claude (or any MCP client) through Zotero's built-in local HTTP API. No API keys, no cloud — it talks to the Zotero desktop app running on your machine.
What it can do
All tools are read-only:
Tool | Purpose |
| Search the library (title/creator/year, or everything) |
| Full metadata for one item |
| Notes, attachments, and annotations of an item |
| Extracted full text of a PDF/attachment |
| Browse the collection tree |
| Items in a collection |
| List/filter tags |
| List saved searches |
| Export items as BibTeX |
| Formatted citation + bibliography in any CSL style |
| Check connectivity |
Related MCP server: zotero-mcp
Prerequisites
Zotero 7 or newer, running.
Enable the local API: Zotero → Settings → Advanced → "Allow other applications on this computer to communicate with Zotero".
uvinstalled.
Verify the API is up:
curl 'http://localhost:23119/api/users/0/items?limit=1'Install & test
cd ~/zotero-mcp
uv sync
uv run pytest # unit tests (no running Zotero needed)Smoke-test against your live library (Zotero must be open):
uv run python -c "from zotero_mcp.client import ZoteroLocalClient as C; import json; print(json.dumps(C().search_items(q='', limit=3), default=str)[:600])"Inspect the tools interactively:
uv run mcp dev src/zotero_mcp/server.pyRegister with Claude Code
claude mcp add zotero -- uv --directory ~/zotero-mcp run zotero-mcpThen ask Claude things like "search my Zotero for papers on diffusion models" or "give me the BibTeX for item ABCD1234".
Configuration
Environment variables (optional):
ZOTERO_BASE_URL— override the API base (defaulthttp://localhost:23119/api). Useful from WSL where Zotero runs on the Windows host, e.g.http://<windows-host-ip>:23119/api.ZOTERO_USER_ID— override the local user id (default0; almost always correct).
Notes
Read-only by design. Zotero's local API does not support writes; this server intentionally mirrors that. Citation/BibTeX formatting is delegated to Zotero itself.
If a tool returns
{"error": "Could not reach the Zotero local API..."}, Zotero is closed or the local-API setting is off.
Available Tools
11 toolszotero_collection_itemsC
List the top-level items in a collection.
Inputs: collection_key (required), limit. Returns: result
zotero_export_bibtexC
Export one or more items as BibTeX (rendered by Zotero).
Inputs: item_keys (required). Returns: result
zotero_export_citationA
Return formatted citation + bibliography HTML for items, in the given CSL style.
Args: item_keys: Items to format. style: A CSL style id, e.g. "apa", "chicago-note-bibliography", "ieee".
Inputs: item_keys (required), style. Returns: result
zotero_get_childrenA
List the notes, attachments, and annotations belonging to an item.
Useful to find the attachment key needed for zotero_get_item_fulltext, or to read a paper's notes.
Inputs: item_key (required). Returns: result
zotero_get_itemA
Get the full metadata for a single item by its key.
Inputs: item_key (required). Returns: object
zotero_get_item_fulltextA
Return the extracted full text for an item.
Accepts either an attachment key directly, or a parent item key (in which case the first attachment with indexed text is used). Full text is only available when Zotero has indexed the attachment.
Inputs: item_key (required). Returns: object
zotero_healthA
Check that the Zotero local API is reachable. Use this to debug connectivity.
Inputs: no named parameters. Returns: object
zotero_list_collectionsA
List collections. Pass a collection key as parent to list its subcollections.
Inputs: parent. Returns: result
zotero_list_tagsA
List tags in the library, optionally filtered by a substring query.
Inputs: limit, query. Returns: result
zotero_saved_searchesA
List the library's saved searches (name + key + conditions).
Inputs: no named parameters. Returns: result
zotero_searchA
Search the Zotero library for top-level items.
Args: query: Free-text search. Matches title/creator/year by default. item_type: Restrict to a Zotero item type, e.g. "journalArticle", "book". tag: Restrict to items carrying this tag. limit: Maximum number of items to return. everything: If true, search all fields and full-text rather than just title/creator/year.
Returns a compact list of items, each with its key (use it with
zotero_get_item / zotero_get_children / zotero_get_item_fulltext).
Inputs: tag, limit, query, item_type, +1 more. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 11 tools
Each tool has a clearly distinct purpose: listing, searching, getting metadata/children/fulltext, exporting in different formats, checking health, and listing collections/tags/saved searches. No overlapping functionality that would cause confusion.
All tools share the 'zotero_' prefix and mostly follow a verb_noun pattern (e.g., zotero_get_item, zotero_search, zotero_list_collections). Minor deviations: 'zotero_collection_items' and 'zotero_saved_searches' are noun phrases rather than explicit verbs, but still clear.
11 tools is well-scoped for a research reference management server. Each tool serves a necessary function without redundancy, covering search, retrieval, listing, export, and health checks.
The tool set provides thorough read capabilities: search, get items/fulltext/children, list collections/tags/saved searches, and export. It lacks create/update/delete operations, but the server appears designed for querying and exporting, so the surface is appropriate for that purpose.
Maintenance
Related MCP Connectors
Zotero MCP server for Claude and ChatGPT: search, citations, safe writes, PDF passages and pages.
Remote MCP server for full read/write access to a Zotero library
Read-only MCP server exposing a user ORANO library to their own AI agent.
Search your AI chat history (ChatGPT, Claude, Codex) from any MCP client. Remote, private, read-only
Related MCP Servers
- AlicenseAqualityDmaintenanceA server that enables MCP clients like Anthropic Claude App to interact with local Zotero libraries, allowing users to search papers, manage notes, and access research materials through natural language.1029Apache 2.0
- FlicenseNot gradedqualityDmaintenanceRead-only MCP server for browsing, searching, and exporting a Zotero library from AI assistants.-
- AlicenseNot gradedqualityDmaintenanceThis MCP server enables AI assistants like Claude to read and search your local Zotero bibliographic database directly from your machine.8MIT
- AlicenseNot gradedqualityBmaintenanceLocal read-only MCP server for Zotero libraries, enabling search, retrieval, and full-text access via the Zotero Web API.MIT