hf-papers-mcp
Enables assistants to search, fetch daily curated lists, retrieve metadata, and generate citations for Hugging Face papers.
README
<p align="center"> <img src="./assets/header.svg" alt="hf-papers-mcp — Hugging Face Papers, inside your assistant" width="100%"> </p>
An MCP server that lets your assistant read Hugging Face Papers — search them, pull the daily curated list, fetch metadata, and generate citations.
Ask "what shipped on speculative decoding this week?" and get real papers with upvote counts and abstracts, instead of whatever was in the model's training data.
Install
Clone it, then point your MCP client (Claude Desktop, Claude Code, or any MCP host) at
server.js:
git clone https://github.com/QEbellavita/hf-papers-mcp
{
"mcpServers": {
"hf-papers": {
"command": "node",
"args": ["/absolute/path/to/hf-papers-mcp/server.js"]
}
}
}
Not on npm — use the clone above rather than npx hf-papers-mcp.
Requires uv. The Python side declares its
dependencies inline (PEP 723), so uv run resolves them on first call — no virtualenv to
create or pip install to remember. Without uv it falls back to bare python3, which
lacks huggingface_hub and will return degraded results rather than pretending
everything is fine.
No API key needed for reads. Set HF_TOKEN only if you want link or claim.
Tools
| Tool | What it does |
|---|---|
hf_papers_search |
Search papers by keyword |
hf_papers_daily |
The daily curated list, optionally for a past date |
hf_papers_info |
Full metadata for an arXiv ID — title, authors, abstract, URLs |
hf_papers_citation |
BibTeX, APA or MLA |
hf_papers_check |
Whether a paper is indexed on HF |
hf_papers_index |
Trigger indexing of an arXiv paper on HF |
Example
> what were the top papers on Hugging Face yesterday?
[1] Domino: Decoupling Causal Modeling from Autoregressive Drafting
in Speculative Decoding
arXiv: 2605.29707 | Upvotes: 152
Bulk digests
Pull a date range in one go:
node bin/daily.js 2026-07-01 2026-07-14
Writes JSON keyed by date to /tmp/, and records the window in
.hf_papers_last_run.json so a bare node bin/daily.js picks up where the last run
finished.
Tests
npm test # hermetic — no network
HF_PAPERS_E2E=1 npm test # adds live API tests
The default suite is offline and covers tool definitions plus a real MCP handshake over stdio. The live tests hit the Hugging Face API and assert that responses come back as parsed JSON — a regression guard, because diagnostics written to stdout used to corrupt the JSON stream and silently degrade structured responses into raw strings.
Notes
Everything runs as a fresh subprocess per call — no daemon, no cached state, nothing
persisted beyond the optional bin/daily.js run marker. Diagnostics go to stderr, since
stdout is the MCP transport and anything else written there breaks the protocol.
Licence
MIT — see LICENSE.
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。