mcp-academic-hub

A unified MCP (Model Context Protocol) server for searching and downloading academic papers and ebooks from 10+ sources.

Sources

Papers

OpenAlex - Open scholarly metadata
ArXiv - Preprints
Unpaywall - Legal Open Access PDFs
CORE - Library Open Access
Sci-Hub - Fallback

Books

Anna's Archive
Z-Library (mirror)
OceanofPDF
PDF Drive
ManyBooks (public domain)
BDeBooks
DigiLibraries
Smashwords

Installation

git clone https://github.com/bockybocky/mcp-academic-hub.git
cd mcp-academic-hub
pip install -r requirements.txt

Configuration

Set environment variables (optional):

export UNPAYWALL_EMAIL="your-email@example.com"  # For Unpaywall API
export CORE_API_KEY="your-core-api-key"           # For CORE API (free at https://core.ac.uk)
export DOWNLOAD_DIR="/path/to/downloads"          # Default download directory

Usage with Claude Code

Add to your Claude Code MCP settings (~/.claude/settings.json):

{
  "mcpServers": {
    "academic-hub": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "server.py"],
      "cwd": "/path/to/mcp-academic-hub"
    }
  }
}

Usage with Gemini CLI

Add to your Gemini MCP settings:

{
  "mcpServers": {
    "academic-hub": {
      "command": "uv",
      "args": ["run", "server.py"],
      "cwd": "/path/to/mcp-academic-hub"
    }
  }
}

Available Tools

Search Tools

| Tool | Description | |------|-------------| | search_papers | Search OpenAlex + ArXiv | | search_ocean_of_pdf | Search OceanofPDF | | search_pdf_drive | Search PDF Drive | | search_manybooks | Search ManyBooks (free/public domain) | | search_bdebooks | Search BDeBooks | | search_digilibraries | Search DigiLibraries | | search_smashwords | Search Smashwords | | search_zlibrary_to | Search Z-Library mirror | | search_annas_archive | Search Anna's Archive |

Download/Link Extraction Tools

| Tool | Description | |------|-------------| | get_pdf_links | Find PDF links for a DOI (Unpaywall + CORE + Sci-Hub) | | download_paper | Download a PDF from a direct URL | | get_ocean_of_pdf_links | Extract download links from OceanofPDF | | get_pdf_drive_links | Extract PDF links from PDF Drive | | get_digilibraries_links | Extract download links from DigiLibraries | | get_zlibrary_to_links | Extract download links from Z-Library | | get_smashwords_links | Extract Smashwords book info |

Requirements

Python >= 3.10
uv (recommended) or pip

License

MIT

MCP Servers