Zotero MCP
Provides tools for searching, reading, and matching Zotero items, managing collections, reading annotations, notes, PDF full text, and citation keys, and optionally importing papers and modifying collections via plan/apply workflow.
Click on "Install 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 MCPsearch my Zotero library for papers on climate change"
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
面向 Windows 与 WSL 的本地 Zotero MCP。读取走 Zotero Local API;需要修改 Zotero 时,先生成计划,再通过 Zotero Web API 执行并回读核验。
交给 Codex 配置
clone 仓库后,在仓库目录对 Codex 发送:
读取 AGENTS.md,按首次配置流程检查我的 Codex 与 Zotero 分别运行在 Windows 还是 WSL。账号注册、浏览器设置和密钥输入由我完成;你负责其余检测、配置与验证。
Codex 会先做只读检查。缺少 Zotero、MinerU 或 SciVerse 账号时,会给出官方入口并暂停;不会索要或显示密钥。
Related MCP server: zotero-grounded-mcp
安装
需要 Python 3.11 或更新版本。
git clone <repository-url> zotero-mcp
cd zotero-mcp
python -m venv .venvWindows PowerShell 安装:
.\.venv\Scripts\Activate.ps1
python -m pip install .WSL/Linux 安装:
source .venv/bin/activate
python -m pip install .运行首次检查:
zotero-mcp setup plan --profile corecore 只要求 Zotero 读取链路。需要 Web API 写入、MinerU、QMD 与 SciVerse 时使用:
zotero-mcp setup plan --profile full完整步骤见 docs/SETUP.md,账号边界见 docs/ACCOUNTS.md。
功能
检索、读取、匹配 Zotero 条目
按 key、唯一名称或完整路径解析 collection
读取批注、笔记、PDF 全文、页面与 Better BibTeX citation key
通过 plan/apply 两阶段工具导入论文、调整 collection、删除 PDF 附件
可选使用 MinerU 批量解析 PDF,并用 QMD 建立全文检索
按 toolset 控制暴露给 Agent 的工具
所有 apply 工具要求精确 key 和明确的 confirm=true。Web API 写入前会重新读取云端状态,写后回读核验。
Toolsets
literature:检索、条目读取、collection 与论文导入review:批注、笔记、全文、MinerU 与页面工具maintenance:collection 调整与 PDF 附件删除all:全部公共工具
默认只启用 literature。生成 Codex 配置:
zotero-mcp setup print-codex-config --toolsets literature,review,maintenance命令只输出配置块,不修改 Codex 配置文件,不包含密钥。
可用 MCP 工具:
zotero_pingzotero_searchzotero_matchzotero_itemzotero_childrenzotero_get_annotationszotero_get_noteszotero_get_citation_keyzotero_collectionszotero_resolve_collectionzotero_item_collectionszotero_collection_itemszotero_web_api_statuszotero_plan_paper_importzotero_apply_paper_importzotero_plan_collection_reconcilezotero_apply_collection_reconcilezotero_plan_pdf_attachment_deletezotero_apply_pdf_attachment_deletezotero_extract_textzotero_mineru_submitzotero_mineru_resultzotero_render_pageszotero_find_figure_pages
可选集成
Zotero Web API:只在写入时需要。
MinerU:外部 PDF 解析服务,需要账号与 Token。
QMD:本地检索工具,无账号要求。
SciVerse:独立 MCP,用于文献检索,需要单独注册与登录。
WebDAV:仅用于 Zotero 附件同步,不是 MCP 依赖。
完整文献工作流保持四段独立边界:SciVerse 检索候选,Zotero 查重并经 plan/apply 导入,MinerU 解析已确认的 PDF,QMD 更新索引并回读核验。任何外部上传或 Zotero 写入都需要单独批准。
配置位置
Windows:
%APPDATA%\zotero-mcp\config.tomlWSL/Linux:
~/.config/zotero-mcp/config.toml
可用 ZOTERO_MCP_CONFIG 指定配置文件,或用 ZOTERO_MCP_CONFIG_DIR 指定配置目录。环境变量与完整配置项见 docs/SETUP.md。
MinerU 批处理
单批预检与提交:
zotero-mineru plan <collection-key> --recursive
zotero-mineru submit-batch <collection-key> --recursive --max-pages 1000 --max-files 20
zotero-mineru collect <batch-id>
zotero-mineru verify <batch-id>MinerU 与 QMD 的有界流水线:
zotero-mineru-qmd <collection-key> --recursive --page-budget 1000 --max-files 20批次状态保存在 MinerU 输出目录的 .jobs 中。ledger 默认是同目录下的 mineru_todo.csv,也可通过配置项 mineru.ledger 或环境变量 ZOTERO_MINERU_LEDGER 指定。
测试
python -m unittest discover -v常见故障见 docs/TROUBLESHOOTING.md。
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityCmaintenanceRead-only MCP server that lets Claude or any MCP client search and retrieve metadata, notes, full text, citations, and BibTeX from your local Zotero library via its built-in API.11MIT
- AlicenseAqualityDmaintenanceAn MCP server that gives any MCP-compatible assistant access to your Zotero reference library, enabling search, citation, bibliography generation, and .docx processing while keeping Zotero as the ground truth for references.91MIT
- AlicenseAqualityAmaintenanceAn MCP server that gives AI assistants complete, safe access to your Zotero library for searching, citing, adding papers, and formatting bibliographies, with local-first privacy.2829323MIT
- Flicense-qualityCmaintenanceMCP server for interacting with a Zotero library via the local API. Enables searching, retrieving, creating, updating, and deleting Zotero items, managing collections and tags, and generating citations.
Related MCP Connectors
The everything Zotero MCP server — Web API v3 + local API, safe writes, citations, search.
Remote MCP server for full read/write access to a Zotero library
MCP-native open-source Notion alternative: read & write pages, databases and kanban boards.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/MurmuredVelpro/My-Zotero-MCP'
If you have feedback or need assistance with the MCP directory API, please join our Discord server