Scopus MCP Server
Scopus MCP Server enables researchers to interact with the Scopus database through natural language. It provides tools for searching documents, fetching abstracts, finding author profiles, and viewing citation overviews.
README
Scopus MCP Server
A local MCP server that wraps the Elsevier Scopus APIs, giving Claude Code four tools:
scopus_search— search Scopus documents by keyword, title, author, affiliation, etc.scopus_get_abstract— fetch a full abstract + metadata by Scopus ID / DOI / EIDscopus_search_authors— find author profiles by name/affiliationscopus_get_author_profile— full author profile (h-index, affiliation, ORCID, etc.) by Scopus Author IDscopus_get_citation_overview— year-by-year citation counts for one or more documents
1. Get an API key
Register at https://dev.elsevier.com/apikey to get a Scopus/Elsevier API key. Some endpoints (especially abstract/citation data) require your key to be used from an IP range or network your institution has registered with Elsevier — if you get 403 errors, that's usually why.
If you also have an institutional token, note it too (optional).
2. Install dependencies
cd "scopus-mcp-server"
pip install -r requirements.txt
3. Set your API key
Set it as an environment variable — never hardcode it in the source or paste it into chat. Pick one option:
Windows PowerShell (current session only):
$env:SCOPUS_API_KEY = "your-key-here"
Windows (persistent, current user):
setx SCOPUS_API_KEY "your-key-here"
(Open a new terminal afterward for setx to take effect.)
Optional institutional token:
setx SCOPUS_INSTTOKEN "your-insttoken-here"
4. Register the server with Claude Code
From an interactive terminal (this step needs claude mcp, which isn't available in
a non-interactive session):
claude mcp add scopus -- python "C:\Users\Hype AMD\OneDrive\Desktop\Claude Cowork\scopus-mcp-server\server.py"
This registers it as a stdio MCP server that Claude Code will launch on demand. Verify
with claude mcp list, then restart/reload Claude Code so it picks up the new server.
5. Try it
Ask Claude something like:
- "Search Scopus for papers on CRISPR gene editing published after 2022"
- "Get the abstract for Scopus ID 85123456789"
- "Find the Scopus author profile for John Smith at MIT"
- "How has paper 85123456789 been cited over the last 5 years?"
Notes
- All tools are read-only (no writes to Scopus).
- Responses default to Markdown; pass
response_format: "json"for structured output. - If you see
Error: SCOPUS_API_KEY environment variable is not set, the server process didn't inherit the env var — set it persistently (setx) and open a fresh terminal / restart Claude Code. - If you see 403 errors, your key likely isn't entitled for that endpoint or needs to be called from your institution's registered network.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。