europepmc-mcp-server
Enables searching and retrieving biomedical literature from Europe PMC, including abstracts, full-text (JATS XML), text-mined annotations, citations, references, and database cross-links, through natural language queries and automated data staging.
README
Europe PMC MCP Server
A Model Context Protocol (MCP) server for Europe PMC — open-access biomedical literature: search across 45M+ abstracts, retrieve OA full-text (JATS XML), text-mined annotations, citations/references, and database cross-links. One of 100+ servers in the Bio MCP monorepo.
Connect
https://europepmc-mcp-server.quentincody.workers.dev/mcp
Local dev runs at http://localhost:8904/mcp (./scripts/dev-servers.sh europepmc).
Tools
europepmc_search— discover endpoints (Code Mode catalog search)europepmc_execute— Code Mode:api.get(...)in a V8 isolate across two upstreams (Articles + Annotations APIs)europepmc_query_data— SQL over large responses auto-staged into per-session SQLiteeuropepmc_get_schema— inspect a staged dataset's schema
Wraps two upstream bases in one isolate (routed in api-adapter.ts): the Articles RESTful API (/search, /{PMCID}/fullTextXML, /{source}/{id}/citations|references|databaseLinks) and the Annotations API (/annotations/annotationsByArticleIds|annotationsByEntity). Every _execute result carries a _meta.citation (Europe PMC / EMBL-EBI).
Key usage notes (also in the catalog notes): pagination is cursorMark (start *, thread nextCursorMark; the page param is ignored); format=json is injected automatically; full text is XML-only and auto-staged; metadata/abstracts are open + commercial-OK, but per-article full-text licenses are mixed (CC-BY vs CC-BY-NC/-ND) — gate any caching/redistribution on the per-article license field (present in resultType=core).
Architecture / maintenance
- Archetype: catalog-only REST + Code Mode (dual upstream). No auth. Rate limit ~10 req/s per IP, shared across Cloudflare egress → be gentle.
- Drift risk: low. Refresh cadence: quarterly catalog review. Known gaps:
supplementaryFiles(binary ZIP) not exposed in v1; standalonetextMinedTerms/fullTextUrlListare 404 upstream (use the Annotations API / thecoreresult's inline fields).
Development
./scripts/dev-servers.sh europepmc
pnpm --filter europepmc-mcp-server run deploy
See docs/adding-mcp-servers.md and the build plan docs/plans/2026-06-26-evidence-genomics-mcp-servers.md.
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。