Zotero MCP 6a001fde

MCP文档知识

提供方 drxaibi

文档知识 未标注 无需鉴权 社区 / 第三方 ClaudeCursor

一个提供全面访问Zotero研究库的MCP服务器,包括元数据、摘要、注释和全文PDF内容。

一句话选型结论

按资料字段整理

Zotero MCP 6a001fde 属于文档知识类 MCP 服务器,传输方式未标注,无需鉴权,可直接接入 Claude、Cursor,没有现成安装命令,需按官方文档接入。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类文档知识 传输方式未标注 鉴权无需鉴权 部署方式未标注
GitHub 星标 0 仓库热度,无公开仓库时显示未标注
工具数 12 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

文档知识
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量12暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

研究库管理TypeScript本地部署学术文献检索PDF文本提取参考文献生成知识管理

安装与配置

未标注 无需鉴权
上游未给出传输方式
  1. 1
    准备运行环境

    按官方文档确认运行环境(常见为 Node.js 或 Python),装好后再执行安装命令。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后客户端会列出约 12 个工具,随便调一个只读工具确认能返回结果。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
文档检索问答

在你的文档库里按问题找答案。

知识库读写

把新资料写进知识库,随用随取。

内容整理

把零散资料归纳成结构化笔记。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 资料标注 12 个工具,覆盖面较广,能满足多种调用。
  • 已适配 Claude、Cursor,换客户端不用重新研究配置。
  • 资料标注无需鉴权,装好就能连。

先想清楚的地方

  • 没有现成安装命令,需要按官方文档自己填启动命令和参数。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在文档知识场景里让模型直接调用外部能力的人。
  • 已经在用 Claude、Cursor 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Zotero MCP 6a001fde当前 12 未标注 未标注 无 2026-05-17
dify 未标注 stdio · 本地进程 本地部署 有 2026-04-11
open-webui 未标注 stdio · 本地进程 本地部署 有 2026-04-11
JeecgBoot 未标注 stdio · 本地进程 本地部署 有 2026-04-11
Cherry Studio 未标注 未标注 未标注 无 2026-05-17
mempalace 未标注 stdio · 本地进程 本地部署 有 2026-04-11

服务器档案

文档知识
字段来自项目仓库与公开索引,中文说明由本站标注
提供方drxaibi上游标注的维护方 drxaibi
分类文档知识本站按上游分类展示
连接方式未标注上游未给出传输方式
鉴权方式无需鉴权远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · 未标注本地安装时用到的包标识与运行时
部署类型未标注上游给出的部署形态
最近核验2026-05-17上游最后核验时间

详细介绍

11.0 KB
来自项目仓库与上游目录的原文,未做删改

Zotero MCP Server

A Model Context Protocol (MCP) server that provides comprehensive access to your Zotero research library. Give LLMs full context about your papers, including metadata, abstracts, notes, annotations, and full-text PDF content.

Features

  • 🔍 Full Library Search: Search items by title, author, tags, or full-text content
  • 📚 Complete Metadata: Access all item fields, creators, tags, and collections
  • 📄 Full-Text Access: Extract and search PDF content
  • ✏️ Annotations & Notes: Access PDF highlights, comments, and user notes
  • 📖 Citations: Generate formatted bibliographies in various styles (APA, MLA, Chicago, etc.)
  • 🔄 Dual Backend: Switch between Web API (remote) and local SQLite database

Quick Start

Prerequisites

  • Node.js >= 18.0.0
  • A Zotero account with API key (Web API mode only — not needed for local mode)

Check if Node.js is installed:```bash
node -v


If not installed or version is below 18, download from nodejs.org (LTS recommended) or use a version manager:```bash
# Windows (winget)
winget install OpenJS.NodeJS.LTS

# macOS (Homebrew)
brew install node

# Linux (apt)
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs

Installation```bash

Clone the repository

git clone https://github.com/drxaibi/zotero-mcp.git
cd zotero-mcp

Install dependencies

npm install

Build the project

npm run build


### Get Your Zotero API Key

1. Go to Zotero API Keys
2. Create a new key with these permissions:
   - ✅ Allow library access
   - ✅ Allow notes access
   - ❌ Allow write access (not needed)
3. Note your **User ID** shown at the top of the keys page
4. Copy the generated API key

### Configuration

Choose your AI assistant and follow the setup instructions:

---

### MCP Configuration

Add the following configuration to your MCP client. Replace the path with your actual `zotero-mcp` location:```jsonc
{
  "mcpServers": {
    "zotero": {
      "command": "node",
      "args": ["C:/path/to/zotero-mcp/dist/index.js"],
      "env": {
        // Web API mode (default)
        "ZOTERO_API_KEY": "your-api-key-here",
        "ZOTERO_USER_ID": "your-user-id-here"

        // Local mode (uncomment below, comment out API_KEY and USER_ID above)
        // "ZOTERO_MODE": "local",
        // "ZOTERO_DATA_DIR": ""  // Leave empty for auto-detection, or set your path
      }
    }
  }
}

> Path format: Use forward slashes (/) or escaped backslashes (\\). Always use absolute paths.
> Local mode: Close Zotero app before using (SQLite file locking).


Note for VS Code + Copilot: Use "servers" instead of "mcpServers" as the root key:```jsonc
{
"servers": {
"zotero": {
"command": "node",
"args": ["C:/path/to/zotero-mcp/dist/index.js"],
"env": {
// Web API mode (default)
"ZOTERO_API_KEY": "your-api-key-here",
"ZOTERO_USER_ID": "your-user-id-here"

    // Local mode (uncomment below, comment out API_KEY and USER_ID above)
    // "ZOTERO_MODE": "local",
    // "ZOTERO_DATA_DIR": ""  // Leave empty for auto-detection, or set your path
  }
}

}
}


### Where to Add Configuration

Client · Config File Location · Root Key
**Claude Desktop** · Windows: `%APPDATA%\Claude\claude_desktop_config.json`
macOS: `~/Library/Application Support/Claude/claude_desktop_config.json` · `mcpServers`
**VS Code + Copilot** · `.vscode/mcp.json` in workspace · `servers`
**VS Code + Claude** · `.vscode/mcp.json` in workspace · `mcpServers`
**Claude Code (CLI)** · Windows: `%USERPROFILE%\.claude\settings.json`
macOS/Linux: `~/.claude/settings.json` · `mcpServers`
**Cursor** · `.cursor/mcp.json` in workspace
or Settings → Features → MCP Servers · `mcpServers`

After adding configuration, restart your client to load the server.

---

### Configuration Options

| Setting | Default | Description |
|---------|---------|-------------|
| `ZOTERO_API_KEY` | *required* | Your Zotero API key |
| `ZOTERO_USER_ID` | *required* | Your Zotero user ID |
| `ZOTERO_MODE` | `web` | `web` (API) or `local` (SQLite) |
| `ZOTERO_GROUP_ID` | *empty* | Group library ID (if accessing group instead of personal) |
| `ZOTERO_DEFAULT_LIMIT` | `25` | Results per query (1-100) |
| `ZOTERO_CACHE_TTL` | `300` | Cache duration in seconds |
| `ZOTERO_MAX_FULLTEXT_LENGTH` | `100000` | Max characters for full-text extraction |
| `ZOTERO_DATA_DIR` | *auto-detected* | Path to Zotero data folder (local mode only) |
| `ZOTERO_CACHE_ENABLED` | `true` | Enable/disable response caching |
| `ZOTERO_EXTRACT_PDF` | `true` | Enable PDF text extraction |

**Full configuration with all options:**```jsonc
{
  "mcpServers": {
    "zotero": {
      "command": "node",
      "args": ["C:/path/to/zotero-mcp/dist/index.js"],
      "env": {
        // Required for Web API mode
        "ZOTERO_API_KEY": "your-api-key-here",
        "ZOTERO_USER_ID": "your-user-id-here",

        // Optional settings
        "ZOTERO_MODE": "web",
        "ZOTERO_GROUP_ID": "",
        "ZOTERO_DEFAULT_LIMIT": "25",
        "ZOTERO_CACHE_ENABLED": "true",
        "ZOTERO_CACHE_TTL": "300",
        "ZOTERO_EXTRACT_PDF": "true",
        "ZOTERO_MAX_FULLTEXT_LENGTH": "100000"

        // Local mode (uncomment and comment out API_KEY/USER_ID above)
        // "ZOTERO_MODE": "local",
        // "ZOTERO_DATA_DIR": ""  // Leave empty for auto-detection, or set your path
      }
    }
  }
}

Local mode notes:
- No API key needed - just set ZOTERO_MODE to local
- Zotero app must be closed when using local mode (SQLite file locking)

Finding your Zotero data directory:

  1. Open Zotero app
  2. Go to Edit → Settings (Windows/Linux) or Zotero → Settings (macOS)
  3. Click Advanced → Files and Folders
  4. Look for Data Directory Location — copy this path

Auto-detection: Leave ZOTERO_DATA_DIR empty ("") or omit it entirely. The server will automatically look for:
- Windows: C:\Users\\Zotero
- macOS: /Users//Zotero
- Linux: /home//Zotero

If auto-detection fails, set the exact path from Zotero settings:```jsonc
"ZOTERO_DATA_DIR": "C:/Users/John/Zotero" // Use YOUR actual path


## Available Tools

| Tool | Description |
|------|-------------|
| `search_library` | Search items by query, tags, type, or collection |
| `get_item` | Get complete item details with all metadata |
| `get_item_fulltext` | Extract full text from PDFs |
| `get_item_notes` | Get all notes attached to an item |
| `get_item_annotations` | Get PDF highlights and comments |
| `list_collections` | List all collections |
| `get_collection_items` | Get items in a collection |
| `list_tags` | List all tags with counts |
| `search_fulltext` | Search across all PDF content |
| `get_bibliography` | Generate formatted citations |
| `get_related_items` | Find related items |
| `get_recent_items` | Get recently added/modified items |

## Available Resources

| Resource | Description |
|----------|-------------|
| `zotero://library/stats` | Library statistics overview |
| `zotero://collections` | Full collection hierarchy |
| `zotero://tags` | All tags with item counts |
| `zotero://recent` | Recently modified items |
| `zotero://schema/item-types` | Supported item types |

## Example Prompts

Once connected, you can ask Claude things like:

- *"Search my Zotero library for papers about machine learning"*
- *"Get the full text of the paper with key ABC123"*
- *"What annotations did I make on this paper?"*
- *"Show me all papers in my 'Literature Review' collection"*
- *"Find papers tagged with 'methodology'"*
- *"Generate APA citations for these three papers"*
- *"What papers did I add this week?"*
- *"Summarize the key findings from my climate change papers"*

## Architecture```
┌─────────────────────────────────────────────────────────────┐
│                     MCP Server (index.ts)                   │
├─────────────────────────────────────────────────────────────┤
│  Tools              │  Resources                            │
│  - search_library   │  - zotero://library/stats             │
│  - get_item         │  - zotero://collections               │
│  - get_fulltext     │  - zotero://tags                      │
│  - ...              │  - ...                                │
├─────────────────────┴───────────────────────────────────────┤
│                    Backend Interface                        │
├─────────────────────────────────────────────────────────────┤
│   Web API Backend          │      Local SQLite Backend      │
│   (api.zotero.org)         │      (zotero.sqlite)           │
└─────────────────────────────────────────────────────────────┘

Troubleshooting

"ZOTERO_API_KEY is required"

Make sure you've set the environment variables in your MCP configuration.

"Could not detect Zotero data directory"

In local mode, set ZOTERO_DATA_DIR to your Zotero data folder path explicitly.

"Database is locked" (Local mode)

Close Zotero before using local mode. Zotero locks the SQLite database while running.

No full-text results

Make sure Zotero has indexed your PDFs. In Zotero: Edit → Preferences → Search → Rebuild Index.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

  1. Fork the repository
  2. Create your feature branch (git checkout -b feature/amazing-feature)
  3. Commit your changes (git commit -m 'Add amazing feature')
  4. Push to the branch (git push origin feature/amazing-feature)
  5. Open a Pull Request

Citation

If you use this project in your research or work, please cite it:bibtex @software{zotero_mcp_2026, author = {drxaibi}, title = {Zotero MCP Server: Model Context Protocol Server for Zotero}, year = {2026}, url = {https://github.com/drxaibi/zotero-mcp}, note = {A Model Context Protocol server providing LLM access to Zotero research libraries} }

Or in text format:
> drxaibi. (2026). Zotero MCP Server: Model Context Protocol Server for Zotero. GitHub. https://github.com/drxaibi/zotero-mcp

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • Zotero - The amazing open-source reference manager
  • Model Context Protocol - The protocol that makes this possible
  • Anthropic - For Claude and the MCP SDK

Related Projects

  • Zotero - The Zotero client
  • pyzotero - Python client for Zotero API
  • MCP Servers - Official MCP server examples

常见问题

6 问
答案由本站依据条目字段整理,不引用上游原文
Zotero MCP 6a001fde 怎么安装?

资料没有给出安装命令,请到官方仓库或文档按说明接入;这类条目多数是生态应用,或尚未发布可安装包。

Zotero MCP 6a001fde 怎么在 Claude、Cursor 里配置?

它的传输方式未标注,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Zotero MCP 6a001fde 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Zotero MCP 6a001fde 需要 API Key 吗?

资料标注为无需认证,装好就能连;如果实际调用时报鉴权错误,按官方文档补上凭据。

Zotero MCP 6a001fde 支持哪些客户端?

资料标注支持 Claude、Cursor;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Zotero MCP 6a001fde 能提供哪些能力?

资料显示约 12 个工具;能力标签包括 TypeScript、Claude、文档处理。具体工具清单以官方文档为准。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-05-17。