MCP 服务器

aigroup-paper-mcp

Academic paper search and retrieval MCP server integrating multiple scholarly platforms into a unified interface. Supports search, fetch, trend analysis, and literature review workflows.

README

aigroup-paper-mcp

Academic paper search and retrieval MCP server integrating multiple scholarly platforms into a unified interface.

Overview

aigroup-paper-mcp provides a unified MCP interface for searching, retrieving, and organizing academic paper metadata across major scholarly sources.

It is designed for:

cross-platform academic paper search
paper metadata retrieval and browsing
literature review assistance
research gap analysis and comparison workflows
integration with Claude Desktop and other MCP-compatible clients

Highlights

12+ academic platforms integrated behind one MCP interface
6 advanced tools for search, fetch, discovery, and trend analysis
3 resource patterns for direct metadata and category access
3 prompt templates for literature-review-style workflows
structured responses, caching, and parallel search support

Supported Sources

The server currently supports sources such as:

arXiv
OpenAlex
PubMed Central (PMC)
Europe PMC
bioRxiv
medRxiv
CORE
Semantic Scholar
Crossref
PubMed
Google Scholar
IACR

Quick Start

Requirements

Node.js >= 18
npm

Install and build locally

git clone https://github.com/jackdark425/aigroup-paper-mcp.git
cd aigroup-paper-mcp
npm install
npm run build
npm start

Run as CLI with npx

npx aigroup-paper-mcp --help
npx aigroup-paper-mcp search "machine learning"
npx aigroup-paper-mcp fetch "2301.00001" --source arxiv

MCP Client Configuration

Claude Desktop / RooCode / compatible MCP clients

{
  "mcpServers": {
    "aigroup-paper-mcp": {
      "command": "npx",
      "args": ["aigroup-paper-mcp"]
    }
  }
}

Tools

`search_papers`

Cross-platform paper search with smart source selection and query optimization.

`fetch_paper`

Fetches detailed metadata for a paper by source and identifier.

`fetch_latest`

Gets the latest papers from a selected source/category.

`list_categories`

Lists supported categories for a given platform.

`advanced_search`

Supports more complex boolean-style academic search queries.

`trend_analysis`

Analyzes topic evolution and publication trends over time.

Resources

paper://{source}/{id}
category://{source}/{category}
search://{query}

Prompt Templates

literature_review
research_gap_analysis
paper_comparison

Environment Variables

Create a .env file if needed:

LOG_LEVEL=info
CACHE_ENABLED=true
CACHE_TTL=3600
MAX_SEARCH_LIMIT=100

Project Structure

aigroup-paper-mcp/
├── src/
├── docs/
├── scripts/
├── package.json
└── README.md

Development

npm run build
npm run test
npm run lint

License & Usage

This project is released under the MIT License.

You may use, copy, modify, merge, publish, distribute, sublicense, and sell copies of this software, including in academic, internal, and commercial contexts, provided that the original copyright notice and license text are preserved.

Please keep in mind:

the software is provided "AS IS", without warranty of any kind
you must retain the relevant copyright and permission notice in copies or substantial portions of the software
downstream usage remains subject to the terms, rate limits, metadata rules, and access restrictions of upstream academic data providers

See the full text in LICENSE.

Acknowledgments

Scholarly Data Ecosystem

Thanks to the academic and open metadata ecosystems that make federated retrieval possible, including arXiv, OpenAlex, PubMed, Crossref, Semantic Scholar, and related services.

MCP Ecosystem

Model Context Protocol SDK
- Repository: https://github.com/modelcontextprotocol/servers
- Role: MCP server integration and tool/resource/prompt model