io.github.piiinpiiins/zotero-mcp-local MCP Server
io.github.piiinpiiins/zotero-mcp-local
Search, read PDFs, and find related papers in your local Zotero library with AI—100% local and read-only.
What is the io.github.piiinpiiins/zotero-mcp-local MCP server?
Zotero-MCP-lite is an MCP server that enables AI assistants like Claude to search, read PDFs, and discover related papers from your local Zotero bibliography database. It operates entirely offline and read-only, integrating directly with Claude Desktop App.
This server bridges your Zotero research library and AI assistants, allowing you to query your papers, extract PDF content, and find connections between documents without sending data to external services. Ideal for researchers and academics who want AI-powered literature analysis while keeping their library private and local.
How to install io.github.piiinpiiins/zotero-mcp-local
Copy-paste configuration for popular MCP clients.
Tools & capabilities
Tools this server exposes to the agent.
Search Zotero library— Search and query your local Zotero bibliography databaseRead PDFs— Extract and read content from PDF files in your Zotero libraryFind related papers— Discover connections and related papers within your library
Use cases
- Search your Zotero library by keywords, authors, or topics to quickly locate relevant papers
- Extract and analyze text from PDF papers to summarize key findings or methodologies
- Find related papers and connections between documents in your research collection
- Use AI to synthesize insights across multiple papers in your library
- Organize and explore your bibliography with natural language queries
io.github.piiinpiiins/zotero-mcp-local MCP server FAQ
It's an MCP server that gives Claude and other AI assistants read-only access to search, read PDFs, and find related papers in your local Zotero library. All processing happens locally on your computer.
The lite version is free. A Pro version (US$30 one-time) is available with additional tools for Connected Papers visualization, cross-paper author insights, and advanced search features.
Install via PyPI (`pip install zotero-mcp-local`), add the server configuration to Claude's `claude_desktop_config.json` file with the correct path to your Zotero-MCP-lite folder, then restart Claude Desktop.
No. The server is 100% local and read-only. It accesses your local Zotero database file (`zotero.sqlite`) and does not require internet or external authentication.
Python 3.10 or higher, the `uv` package manager, and a local Zotero folder containing `zotero.sqlite`. Installation guides are provided for both Mac and Windows.
Yes. You can specify a custom path to your `zotero.sqlite` file by adding a `ZOTERO_DB_PATH` environment variable in the MCP server configuration.
README (reference)
Source of truth, from the repository.
Zotero-MCP-lite 快速設定
這是一個提供給支援 Model Context Protocol (MCP) 客戶端使用的伺服器軟體,讓 AI 能夠具備讀取與搜尋您本機 Zotero 書目資料庫的技能。目前主流多搭配 Claude Desktop App 使用。
前置需求(Mac)
- Python >= 3.10:建議透過 Homebrew 安裝 (
brew install python),或至 Python 官網 下載安裝檔。 - uv 套件管理工具:打開終端機輸入
curl -LsSf https://astral.sh/uv/install.sh | sh或brew install uv - Zotero 資料夾(含
zotero.sqlite)已放置在本機電腦上
Windows 使用者請改看 windows_install_guide.md,裡面有完整的前置需求、安裝步驟與除錯說明。
安裝步驟(Mac)
Windows 使用者請改看 windows_install_guide.md。
-
把整個專案資料夾複製到 Mac 任意位置(從 GitHub 下載 ZIP 解壓縮的話,資料夾名稱通常是
Zotero-MCP-lite-main,下面的範例都以它為準) -
打開終端機 (Terminal),在資料夾內執行:
uv sync -
修改
.mcp.json裡的路徑:- 把
/PATH/TO/Zotero-MCP-lite改成實際路徑
- 把
-
如果 Zotero 資料夾不在預設位置
~/Zotero/,請在 Claude 的 MCP 設定中加入env欄位:{ "mcpServers": { "zotero": { "command": "uv", "args": [ "--directory", "/PATH/TO/Zotero-MCP-lite", "run", "python", "-m", "src.zotero_mcp.server" ], "env": { "ZOTERO_DB_PATH": "/your/path/to/zotero.sqlite" } } } }不要用
export ZOTERO_DB_PATH=...來設定:在終端機下的export只對那個視窗有效,從 Launchpad 或 Dock 開啟的 Claude Desktop 是另一個獨立的程序,讀不到。要讓 Claude 認得路徑,只能寫在設定檔的env欄位裡。(若只是想用下面的「驗證」指令在終端機測試,才需要export。)
如何連接到 Claude Desktop(Mac)
Windows 使用者請改看 windows_install_guide.md。
完成上方的「前置需求」與「安裝步驟」後,透過以下步驟將程式連接至 Claude:
-
開啟 Claude Desktop 的設定檔
claude_desktop_config.json(二擇一):方法 A — 從 Claude Desktop App 內開啟(推薦):
- 開啟 Claude Desktop App
- 點擊左上角選單 Claude → Settings
- 點選左側的 Developer
- 點擊 Edit Config 按鈕,系統會自動用文字編輯器打開
claude_desktop_config.json
方法 B — 手動開啟檔案:打開 Finder,按
Cmd + Shift + G,貼上~/Library/Application Support/Claude/,找到claude_desktop_config.json並用文字編輯器開啟。若檔案不存在,請自行新建。 -
將本資料夾中
.mcp.json的內容複製並合併至claude_desktop_config.json裡面 -
修改路徑:務必將剛貼上內容中的
/PATH/TO/Zotero-MCP-lite更改為本專案資料夾的實際絕對路徑例如,若您將資料夾放在桌面,路徑為
/Users/huang/Desktop/Zotero-MCP-lite-main,則claude_desktop_config.json應修改為:{ "mcpServers": { "zotero": { "command": "uv", "args": [ "--directory", "/Users/huang/Desktop/Zotero-MCP-lite-main", "run", "python", "-m", "src.zotero_mcp.server" ] } } }注意:如果您的
claude_desktop_config.json裡已有其他設定(如coworkScheduledTasksEnabled等),請確保mcpServers與它們平行放置,不要把其他設定放進mcpServers裡面。如果您在「安裝步驟」第 4 步加過
env(Zotero 不在預設位置):上面這份範例沒有env,請不要拿它整段覆蓋掉您的設定,把"env": { "ZOTERO_DB_PATH": "..." }那段保留在"args"後面即可。 -
完全重啟 Claude:儲存設定檔後,用
Cmd + Q完全結束 Claude Desktop(關掉視窗不算),再重新開啟。點開輸入框旁的「搜尋與工具」選單,看到zotero即表示伺服器啟動成功!(舊版 Claude Desktop 是在右下角顯示 🔌 或鐵鎚圖示。)
驗證
先 cd 到專案資料夾,再執行:
uv run python -m src.zotero_mcp.server
怎麼判斷成功:這個伺服器是靠標準輸入輸出溝通的,正常啟動時不會印出任何訊息,游標會停在那裡等待輸入。只要它沒有馬上跳回提示字元、也沒有噴出錯誤,就表示啟動成功,按 Ctrl + C 結束即可。反之,如果指令一秒內就結束並跳回提示字元,代表伺服器沒能啟動。
可能會看到一行
RuntimeWarning: 'src.zotero_mcp.server' found in sys.modules...,這是 Python 的提醒訊息,不影響運作,可以忽略。
除錯
Windows 常見問題請看 windows_install_guide.md 的除錯章節。
Mac 系統常見問題
- 找不到 Zotero 資料庫 (
zotero.sqlite)- 問題:不知道 Zotero 資料庫存放在哪裡,無法確定路徑。
- 解法:預設路徑通常為
~/Zotero/zotero.sqlite。如果找不到,請打開終端機 (Terminal) 輸入以下指令搜尋:- Spotlight 快速搜尋(推薦):
mdfind -name "zotero.sqlite" - 個人目錄深度搜尋:
find ~ -name "zotero.sqlite" 2>/dev/null
- Spotlight 快速搜尋(推薦):
進階版(Pro)
需要更完整的研究分析工具嗎?Pro 版多了四個工具:Connected Papers 關聯圖與視覺化、跨論文整理作者自述的研究限制與未來方向、跨論文關鍵段落搜尋(附頁碼)。
👉 Zotero-MCP Pro(US$30 一次買斷,含中英安裝指南與診斷工具):https://3718181853007.gumroad.com/l/calm-study-partner
Related MCP servers
Create decks, quizzes, matching exercises, and fill-in-the-blank texts on Cuttab via AI.

Google Ads MCP server — query campaigns, keywords, assets & more via natural language

io.github.pijusz/mcp-gsc
Google Search Console MCP server — query performance, sitemaps, indexing & more

io.github.pijusz/mcp-mailtrap
MCP server for the Mailtrap email platform — analytics, sending, contacts, sandbox

io.github.pijusz/mcp-redtrack
MCP server for RedTrack affiliate tracking API

io.github.pijusz/mcp-sanity-images
MCP server for uploading local images to Sanity CMS