Paper Download

Provides multi-source academic paper downloads with intelligent routing and metadata access.
  • python

1

GitHub Stars

python

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "oxidane-bot-paper-download-mcp": {
      "command": "uvx",
      "args": [
        "paper-download-mcp"
      ],
      "env": {
        "PAPER_DOWNLOAD_EMAIL": "your-email@university.edu",
        "PAPER_DOWNLOAD_OUTPUT_DIR": "/path/to/papers"
      }
    }
  }
}

You run a Paper Download MCP Server to fetch academic papers from multiple sources with intelligent routing, automatic fallbacks, and clean filenames. It lets you request a paper by DOI or URL, batch-download multiple papers, and inspect metadata without downloading the PDF, all through MCP-compatible clients.

How to use

Use a MCP client to connect to the Paper Download MCP Server and perform three core actions: download a single paper, download several papers in sequence, or retrieve metadata. The server handles source selection automatically, prioritizing open access options when available and gracefully falling back to other sources if needed. You get a Markdown summary of results, or a JSON payload when requesting metadata.

How to install

Prerequisites: you need a runtime environment that can run MCP servers and a client that can talk the MCP protocol.

Option A — Use a managed environment manager (recommended) to run an MCP server from configuration:

{
  "mcpServers": {
    "paper-download": {
      "command": "uvx",
      "args": ["paper-download-mcp"],
      "env": {
        "PAPER_DOWNLOAD_EMAIL": "your-email@university.edu",
        "PAPER_DOWNLOAD_OUTPUT_DIR": "/path/to/papers"
      }
    }
  }
}

Additional setup steps

If you prefer a local development flow, you can run the server directly after syncing dependencies and then starting the Python module that provides the MCP server.

git clone <repository-url>
cd paper-download-mcp
uv sync
uv run python -m paper_download_mcp.server

Configuration and environment

Key environment variables you may need to set:

PAPER_DOWNLOAD_EMAIL=your-email@university.edu
PAPER_DOWNLOAD_OUTPUT_DIR=/path/to/papers

Claude Desktop setup

If you integrate with Claude Desktop, add the MCP server configuration to your Claude Desktop setup so it can launch paper downloads directly from your workspace.

{
  "mcpServers": {
    "paper-download": {
      "command": "uvx",
      "args": ["paper-download-mcp"],
      "env": {
        "PAPER_DOWNLOAD_EMAIL": "your-email@university.edu",
        "PAPER_DOWNLOAD_OUTPUT_DIR": "/path/to/papers"
      }
    }
  }
}

Troubleshooting

If you see an error about the email requirement, ensure PAPER_DOWNLOAD_EMAIL is set in your environment or in your MCP launcher config.

If a paper cannot be found, verify the DOI or URL, then try metadata lookup first to check availability before attempting a download.

Tools and capabilities

The server exposes three tools to perform common tasks: paper_download to download a single paper by DOI or URL, paper_batch_download to download multiple papers with progress reporting, and paper_metadata to fetch paper details without downloading the PDF.

Available tools

paper_download

Download a single academic paper by DOI or URL. Returns a Markdown summary with download details including file path, size, source, and timing, or an actionable error message if the download fails.

paper_batch_download

Download multiple papers sequentially with progress reporting. Returns a Markdown summary, a list of successful downloads, and a list of failed downloads with errors. Includes a 2-second delay between downloads for rate limiting.

paper_metadata

Retrieve paper metadata without downloading the PDF. Returns details such as DOI, title, year, authors, journal, open access status, and available download sources.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational