ArXiv MCP Server

@blazickjp

代码仓库

基本信息

分类

数据与分析

许可证

Apache-2.0

传输方式

stdio

作者

blazickjp

提交者

zsky1

关于 ArXiv MCP Server

A Model Context Protocol server for searching and analyzing arXiv papers

#ai #gpt #claude-ai #mcp-server

ArXiv MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "arxiv-mcp-server": {
      "command": "uv",
      "args": [
        "tool",
        "run",
        "arxiv-mcp-server",
        "--storage-path",
        "/path/to/paper/storage"
      ]
    }
  }
}

ArXiv MCP Server 工具

Search for papers on arXiv with advanced filtering and query optimization. QUERY CONSTRUCTION GUIDELINES: - Use QUOTED PHRASES for exact matches: "multi-agent systems", "neural networks", "machine learning" - Combine related concepts with OR: "AI agents" OR "software agents" OR "intelligent agents" - Use field-specific searches for precision: - ti:"exact title phrase" - search in titles only - au:"author name" - search by author - abs:"keyword" - search in abstracts only - Use ANDNOT to exclude unwanted results: "machine learning" ANDNOT "survey" - For best results, use 2-4 core concepts rather than long keyword lists ADVANCED SEARCH PATTERNS: - Field + phrase: ti:"transformer architecture" for papers with exact title phrase - Multiple fields: au:"Smith" AND ti:"quantum" for author Smith's quantum papers - Exclusions: "deep learning" ANDNOT ("survey" OR "review") to exclude survey papers - Broad + narrow: "artificial intelligence" AND (robotics OR "computer vision") CATEGORY FILTERING (highly recommended for relevance): - cs.AI: Artificial Intelligence - cs.MA: Multi-Agent Systems - cs.LG: Machine Learning - cs.CL: Computation and Language (NLP) - cs.CV: Computer Vision - cs.RO: Robotics - cs.HC: Human-Computer Interaction - cs.CR: Cryptography and Security - cs.DB: Databases EXAMPLES OF EFFECTIVE QUERIES: - ti:"reinforcement learning" with categories: ["cs.LG", "cs.AI"] - for RL papers by title - au:"Hinton" AND "deep learning" with categories: ["cs.LG"] - for Hinton's deep learning work - "multi-agent" ANDNOT "survey" with categories: ["cs.MA"] - exclude survey papers - abs:"transformer" AND ti:"attention" with categories: ["cs.CL"] - attention papers with transformer abstracts DATE FILTERING: Use YYYY-MM-DD format for historical research: - date_to: "2015-12-31" - for foundational/classic work (pre-2016) - date_from: "2020-01-01" - for recent developments (post-2020) - Both together for specific time periods RESULT QUALITY: Results sorted by RELEVANCE (most relevant papers first), not just newest papers. This ensures you get the most pertinent results regardless of publication date. TIPS FOR FOUNDATIONAL RESEARCH: - Use date_to: "2010-12-31" to find classic papers on BDI, SOAR, ACT-R - Combine with field searches: ti:"BDI" AND abs:"belief desire intention" - Try author searches: au:"Rao" AND "BDI" for Anand Rao's foundational BDI work

Download a paper and create a resource for it

List all existing papers available as resources

Read the full content of a stored paper in markdown format

概览

What is ArXiv MCP Server?

The ArXiv MCP Server provides a bridge between AI assistants and arXiv’s research repository through the Model Context Protocol (MCP). It allows AI models to search for papers and access their content in a programmatic way. The server is for anyone who wants to integrate arXiv paper discovery and retrieval into AI‑powered workflows.

How to use ArXiv MCP Server?

Install the server via Smithery (npx -y @smithery/cli install arxiv-mcp-server --client claude), using the one‑click .mcpb bundle for Claude Desktop on macOS, or manually with uv tool install arxiv-mcp-server. Then add the appropriate MCP configuration (stdio or HTTP) to your client’s config file. The server exposes tools such as search_papers, download_paper, read_paper, list_papers, and semantic_search.

Key features of ArXiv MCP Server

Search arXiv with filters for categories, date ranges, and boolean operators.
Download paper content (HTML first, with PDF fallback).
Read paper content with paging support for large documents.
List all locally downloaded papers.
Local storage for faster repeated access.
Pre‑built research prompts for paper analysis.

Use cases of ArXiv MCP Server

Automate literature searches across arXiv categories and date ranges.
Download and analyze research papers within an AI assistant workflow.
Build a local library of papers for offline reading and semantic search.

常见问题

How do I install ArXiv MCP Server?

Install via Smithery, Claude Desktop .mcpb bundle, or manually with `uv tool install arxiv-mcp-server`. Python 3.11+ must be available.

What are the Python requirements?

The server requires Python 3.11 or newer. For PDF‑only older papers, install the extra with `uv tool install 'arxiv-mcp-server[pdf]'`.

Is paper content safe to feed into AI assistants?

No. arXiv papers are untrusted user‑generated content and may contain prompt injection attacks. Treat all paper text as potentially adversarial, especially in agentic pipelines.

How does the server handle rate limits?

The server enforces arXiv’s 3‑second rate limit automatically. If rate‑limited, wait 60 seconds before retrying.

Can I run the server over HTTP instead of stdio?

Yes. Set the environment variables `TRANSPORT=http`, `HOST`, and `PORT` (e.g., `TRANSPORT=http HOST=127.0.0.1 PORT=8080 arxiv-mcp-server --storage-path /path/to/papers`) and use a Streamable‑HTTP–capable MCP client configured with the server’s URL.

数据与分析分类下的更多 MCP 服务器

Altss Public Profiles MCP

Altss

Anonymous, read-only search and reads of published Altss company-profile summaries, with canonical citations. Hosted HTTP and npm stdio connections; no private app data or account access.

Bounce Watch - Company Signal Intelligence

bouncewatch

Bounce Watch tells you what just changed at a company, and when. Who raised money. Who hired a senior person. Who opened an office, won a customer, or announced a partnership. Every event carries its own date, so you ca

Subtext

Subtext by Fullstory

Session replay, built for agents. Subtext is agentic session review: it captures production sessions of your app and connects them to your coding agent — Claude Code, Cursor, Codex, Devin, your own harness — so it can

LeadMarina

Verified local-business leads: search any niche in any city, query your lead library, and export to Close, GoHighLevel, Google Sheets, or emailed files — every email SMTP-checked, every phone verified with line type and

LeadBoxer

Company and contact enrichment: identify the company behind an IP, get firmographics for a domain, and profile a work email.