search_papersLABSCO SUMMARY
Two tools, and they are the two you need.
Search and fetch is the entire surface, which keeps it out of the way: the assistant finds papers on the real index instead of recalling titles, and pulls the record for the one you name.
What it is
A small Python MCP server for arXiv, installed from PyPI. Two tools: search by query with a result limit, and fetch detailed information for one paper by its arXiv id.
What you get
- Papers searched by query, returning as many results as you ask for
- Detailed information for one paper, by arXiv id
Requirements
Python, with pip install arxiv-search-mcp in the environment your agent runs in. Nothing to authenticate — arXiv search is public.
Setup effort
One command — pip install arxiv-search-mcp
USE CASES
01
Checking whether somebody has already published this
Search the index by the terms you would use, before assuming the idea is new.
02
Getting the real record behind a citation
Fetch the paper by its arXiv id, rather than trusting a remembered title and year.
03
Building a short reading list on a topic
Search broadly, then pull the details for the handful worth opening.
2 tools
Listed in project docs. Your client may show a different list.
get_paperBEFORE YOU INSTALL
What to know first
Tools you get
2 tools named in the project's documentation.
Requires
python on PATH
INSTALLATION
Connect your client
Add the block to your client, or run the command directly.
Server setup
Configuration file
JSON
{
"mcpServers": {
"arxiv": {
"command": "python",
"args": [
"-m",
"arxiv_mcp"
]
}
}
}pip
TERMINAL
pip install arxiv-search-mcp
