Search Papers
Let your agent query arXiv for papers while you scope technical ideas, validate approaches, or gather citations for docs.
Overview
search-papers is a MCP server for the Idea phase that lets agents search arXiv preprints through a stdio npm package.
What is this MCP server?
- Dedicated MCP server for searching arXiv preprints from agent workflows
- npm package search-papers v0.1.1 with stdio transport
- Lightweight research bridge—no custom arXiv API client code in your repo
- Fits early journey research alongside competitive and audience discovery
- Published on npm registry with GitHub source for audit
- Package version 0.1.1
- Transport: stdio
- Registry: npm search-papers
What problem does it solve?
You lose momentum context-switching to arXiv.org in a browser whenever your agent needs recent papers on a technical bet.
Who is it for?
Solo builders and indie ML hackers who want the simplest arXiv search hook inside MCP during early research.
Skip if: Teams needing PubMed, IEEE, paid journal access, PDF full-text mining, or reference managers in one tool.
What do I get? / Deliverables
After registering the stdio server, your agent can return arXiv search results inline while you research and narrow what to build.
- arXiv search results returned to the agent conversation
- Faster technical literature checks during ideation
- Reduced manual browsing on arxiv.org
Recommended MCP Servers
Journey fit
Literature search belongs on the idea-phase shelf when solo builders explore feasibility before committing to a full build. search-papers is a focused stdio MCP integration for arXiv discovery, not a general web browser or note-taking skill.
How it compares
Single-source arXiv MCP search, not a multi-database research assistant or browser automation suite.
Common Questions / FAQ
Who is search-papers for?
Builders and researchers using MCP agents who routinely need arXiv listings while exploring ideas or validating technical directions.
When should I use search-papers?
Use it in idea and early validate research when you want paper titles and metadata from arXiv without manual site searches.
How do I add search-papers to my agent?
Add the npm MCP server search-papers with stdio transport in your client configuration per the package registry entry.