open-articles-mcp
v1.0.0
Published
MCP server for fetching research papers by DOI or URL. Extracts full text from scientific articles.
Downloads
222
Maintainers
Readme
Open Articles MCP
MCP (Model Context Protocol) server for fetching full text of research papers by DOI or URL.
Features
- Fetch research papers by DOI
- Fetch papers by URL (arXiv, PubMed, publisher links)
- Extract full text from PDFs
- Works with Cursor IDE and other MCP-compatible tools
Installation
For Cursor IDE
Add to your Cursor settings (~/.cursor/mcp.json or project .cursor/mcp.json):
{
"mcpServers": {
"open-articles-mcp": {
"command": "npx",
"args": ["-y", "open-articles-mcp"]
}
}
}Custom Server URL
If you're running your own server instance:
{
"mcpServers": {
"open-articles-mcp": {
"command": "npx",
"args": ["-y", "open-articles-mcp"],
"env": {
"MCP_SERVER_URL": "https://your-server.com"
}
}
}
}Usage
Once configured, you can ask your AI assistant to:
- "Fetch the paper with DOI 10.1038/nature12373"
- "Get the full text of https://arxiv.org/abs/2301.00001"
- "Download and analyze the paper at https://doi.org/10.1234/example"
Available Tools
fetch_research_paper_by_doi
Downloads full PDF version of a research paper by DOI.
Parameters:
doi(string, required): Paper DOI (e.g., "10.1176/ajp.120.6.571")
fetch_research_paper_by_url
Downloads full PDF version of a research paper by URL.
Parameters:
url(string, required): URL to the research paper
Supported URL formats:
- Direct DOI URLs:
https://doi.org/10.1234/example - arXiv:
https://arxiv.org/abs/2301.00001 - PubMed:
https://pubmed.ncbi.nlm.nih.gov/12345678 - Publisher links with DOI in URL
Development
# Clone repository
git clone https://github.com/your-username/open-articles-mcp
cd open-articles-mcp/open-articles-mcp
# Install dependencies
npm install
# Run locally
npm startEnvironment Variables
MCP_SERVER_URL: URL of the backend server (default: https://open-articles-mcp.onrender.com)
License
MIT
