Skip to main content
Glama
PaperMCP

PaperMCP

by telagod

PaperMCP

npm version License TypeScript MCP

Search and access academic papers from 23+ sources through Model Context Protocol

English | 简体中文


❌ Without PaperMCP

  • Manually visiting multiple academic databases

  • Copy-pasting DOIs and paper IDs between tools

  • Switching contexts to download PDFs

  • No unified interface for paper search

Related MCP server: Paper Search MCP

✅ With PaperMCP

  • Single interface for 23+ academic sources

  • Direct access from Claude and other AI assistants

  • Automatic PDF download and text extraction

  • Plugin system for sensitive sources


🛠️ Installation

Add to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Using npm global install

npm install -g @telagod/papermcp
{
  "mcpServers": {
    "papermcp": {
      "command": "papermcp"
    }
  }
}

With API keys

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"],
      "env": {
        "SEMANTIC_SCHOLAR_API_KEY": "YOUR_KEY",
        "WOS_API_KEY": "YOUR_KEY",
        "SCOPUS_API_KEY": "YOUR_KEY"
      }
    }
  }
}

Add to Cline MCP settings:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to ~/.config/zed/settings.json:

{
  "context_servers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to Continue config:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to Cody settings:

{
  "cody.experimental.mcp": {
    "servers": {
      "papermcp": {
        "command": "npx",
        "args": ["-y", "@telagod/papermcp"]
      }
    }
  }
}

Add to Cursor MCP settings:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to Windsurf config:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to Roo Cline settings:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to OpenHands config:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to Void settings:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Use the claude mcp add command:

claude mcp add papermcp -- npx -y @telagod/papermcp

Or manually add to ~/.claude.json:

{
  "projects": {
    "/your/project/path": {
      "mcpServers": {
        "papermcp": {
          "command": "npx",
          "args": ["-y", "@telagod/papermcp"]
        }
      }
    }
  }
}

Add to ~/.config/gemini-cli/settings.json:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

Add to ~/.codex/config.toml:

[mcp_servers.papermcp]
command = "npx"
args = ["-y", "@telagod/papermcp"]

Add to .opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "papermcp": {
      "type": "local",
      "command": ["npx", "-y", "@telagod/papermcp"],
      "enabled": true
    }
  }
}

Add to Crush config:

{
  "mcpServers": {
    "papermcp": {
      "command": "npx",
      "args": ["-y", "@telagod/papermcp"]
    }
  }
}

📚 Supported Platforms

Core Platforms (16)

Platform

ID

API Key

Features

arXiv

arxiv

❌

Search, Download, Read

PubMed

pubmed

❌

Search, Lookup

PubMed Central

pmc

❌

Search, Download, Read

bioRxiv

biorxiv

❌

Search

medRxiv

medrxiv

❌

Search

Google Scholar

google-scholar

❌

Search

IACR ePrint

iacr

❌

Search, Download

Semantic Scholar

semantic

⚠️

Search, Lookup

CrossRef

crossref

❌

Search, Lookup

ACM Digital Library

acm

❌

Search

Web of Science

wos

✅

Search

Scopus

scopus

✅

Search

JSTOR

jstor

❌

Search

ResearchGate

researchgate

❌

Search

CORE

core

✅

Search

Microsoft Academic

microsoft-academic

✅

Search

Optional Plugins (7)

Enable via environment variables:

PLUGIN_SCI_HUB=true
PLUGIN_LIBGEN=true
PLUGIN_UNPAYWALL=true
PLUGIN_OA_BUTTON=true
PLUGIN_SCIENCE_DIRECT=true
PLUGIN_SPRINGER_LINK=true
PLUGIN_IEEE_XPLORE=true
WARNING

Sci-Hub and LibGen plugins are disabled by default. Use responsibly and comply with local regulations.


🚀 Usage

Available Tools (4 Unified Tools)

recommend_platforms

Get platform recommendations based on field

{
  "query": "transformer neural network",
  "field": "computer-science"  // biomedical, physics, mathematics, cryptography, open-access, general
}

search_papers

Search for papers on specified platform

{
  "platform": "arxiv",  // Use recommend_platforms to get suggestions
  "query": "machine learning",
  "limit": 10
}

download_paper

Download paper PDF from platform

{
  "platform": "arxiv",
  "id": "2301.00001",
  "dir": "/path/to/save"
}

read_paper

Extract text from paper PDF

{
  "platform": "pmc",
  "id": "PMC8123456",
  "dir": "/path/to/pdfs"
}

⚙️ Configuration

Environment Variables

# Optional API keys for enhanced features
SEMANTIC_SCHOLAR_API_KEY=your-key
WOS_API_KEY=your-key
SCOPUS_API_KEY=your-key
CORE_API_KEY=your-key

# Plugin toggles (default: false)
PLUGIN_SCI_HUB=true              # No config needed
PLUGIN_LIBGEN=true               # No config needed
PLUGIN_UNPAYWALL=true            # Requires UNPAYWALL_EMAIL
PLUGIN_OA_BUTTON=true            # Optional OA_BUTTON_API_KEY
PLUGIN_SCIENCE_DIRECT=true       # Requires ELSEVIER_API_KEY
PLUGIN_SPRINGER_LINK=true        # No config needed
PLUGIN_IEEE_XPLORE=true          # No config needed

# Plugin API keys
UNPAYWALL_EMAIL=your@email.com           # Required for Unpaywall
ELSEVIER_API_KEY=your-key                # Required for ScienceDirect
OA_BUTTON_API_KEY=your-key               # Optional for Open Access Button

Getting API Keys

Platform

Link

Notes

Semantic Scholar

Apply

Free, increases rate limit

Web of Science

Apply

Requires institutional subscription

Scopus

Apply

Requires institutional subscription

CORE

Apply

Free

Unpaywall

Any email

No registration needed

ScienceDirect

Apply

Requires institutional subscription

Open Access Button

Apply

Optional, free


🔧 Development

Install from source

git clone https://github.com/telagod/papermcp.git
cd papermcp/ts
npm install
npm run build
npm run dev

Project Structure

ts/
├── src/
│   ├── core/           # Core types and config
│   ├── platforms/      # Platform adapters
│   ├── plugins/        # Optional plugins
│   ├── services/       # Registry and tools
│   ├── server/         # MCP server
│   └── utils/          # HTTP, logging
└── dist/               # Compiled output

Adding a Platform

import { BasePlatformAdapter } from './baseAdapter.js';
import { addAdapterFactory } from './index.js';

class MyAdapter extends BasePlatformAdapter {
  constructor() {
    super('my-platform');
  }

  async search(query: SearchQuery): Promise<SearchResult> {
    // Implementation
  }
}

addAdapterFactory(() => new MyAdapter());

📄 License

MIT License - see LICENSE for details.


🙏 Acknowledgments

Built with Model Context Protocol and MCP TypeScript SDK


⬆ back to top

Made with ❤️ for researchers

Available Tools

4 tools
download_paper下载论文PDFC

从指定平台下载论文PDF文件

Inputs: id (required), platform (required), dir. Returns: id, path, cached, source

read_paper读取论文内容B

读取指定平台论文的PDF文本内容

Inputs: id (required), platform (required), dir. Returns: id, text, source, statistics

recommend_platforms推荐学术平台A

根据查询意图和学科领域推荐最合适的学术搜索平台。返回按优先级排序的平台列表及其特点。

Inputs: query (required), field. Returns: field, recommendations

search_papers搜索学术论文B

在指定平台搜索学术论文。使用 recommend_platforms 工具获取推荐平台。

Inputs: query (required), platform (required), year, limit, +2 more. Returns: items, source, nextCursor

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 4 tools

    Disambiguation5/5

    Each tool has a distinct role in the workflow: recommend_platforms suggests sources, search_papers finds papers, download_paper retrieves files, and read_paper extracts content. There is no overlap or confusion between these tools.

    Naming Consistency5/5

    All tool names follow the verb_noun pattern with consistent snake_case formatting (recommend_platforms, search_papers, download_paper, read_paper). The naming is uniform and predictable.

    Tool Count5/5

    With only 4 tools, the set is tightly scoped and each tool addresses a necessary step in the paper search and reading pipeline. This is an appropriate size for the server's purpose.

    Completeness5/5

    The tool set covers the complete workflow from platform recommendation, to search, download, and reading. There are no obvious gaps in the core functionality a user would expect from a paper management server.

    Maintenance

    ActivityInactive
    ResponsivenessLimited data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      C
      maintenance
      Provides paper search, PDF link retrieval, and content parsing for arXiv research papers, enabling seamless access to scientific literature through the Model Context Protocol.
      4
      38 npm
      24
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables searching and downloading academic papers from multiple sources including arXiv, PubMed, bioRxiv, Google Scholar, and Semantic Scholar. Provides standardized tools compatible with OpenAI Deep Research and ChatGPT connectors.
      15
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables AI assistants to search for academic papers by DOI, title, or keywords and download full-text PDFs from Sci-Hub. It provides a programmatic interface for accessing metadata and scientific literature through the Model Context Protocol.
      189
      -
    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides real-time access to over 200 million scientific papers and full-text extraction from major academic sources including arXiv, OpenAlex, and PubMed Central. It enables users to search, fetch metadata, and analyze citations across multiple research disciplines through a unified Model Context Protocol interface.
      2,478 npm
      63
      MIT