The arXiv Search Integration for Claude is a Model Context Protocol (MCP) server designed to facilitate the search and retrieval of academic papers from arXiv. This server enables users to access detailed information about academic papers, including metadata, summaries, and PDF links, directly through MCP tools.
arxiv://paper/{id}
URIs.search_papers
tool allows users to search for academic papers on arXiv.query
(e.g., 'machine learning', 'quantum physics').max_results
, sort_by
, and start
for pagination.To get started, install the necessary dependencies:
npm install
Build the server using:
npm run build
For development with auto-rebuild, use:
npm run watch
To use the server with Claude Desktop, add the following configuration:
MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json
{
"mcpServers": {
"arxiv-search": {
"command": "node",
"args": ["/path/to/arxiv-search/build/index.js"],
"disabled": false,
"alwaysAllow": []
}
}
}
Add the following configuration to: ~/Library/Application Support/Code/User/globalStorage/rooveterinaryinc.roo-cline/settings/cline_mcp_settings.json
Debugging MCP servers can be challenging due to stdio communication. Use the MCP Inspector for easier debugging:
npm run inspector
<use_mcp_tool>
<server_name>arxiv-search</server_name>
<tool_name>search_papers</tool_name>
<arguments>
{
"query": "machine learning",
"max_results": 5,
"sort_by": "relevance"
}
</arguments>
</use_mcp_tool>
query
: Search query (required).max_results
: Maximum number of results to return (default: 10, max: 100).sort_by
: Sort method ('relevance', 'lastUpdatedDate', 'submittedDate').start
: Starting index for results (for pagination).This project is licensed under the MIT License.
For a video demonstration, visit this link.
For further inquiries, contact the developer via WeChat: stoeng
.