Hosted MCP server · #155 most used on Smithery

ArXiv Scout

shawnnygoh/arxiv-scout

Search and retrieve academic papers directly from arXiv with advanced query capabilities. Extract full text from PDFs to generate summaries, literature reviews, and side-by-side comparisons. Track citations and references to discover related research and map out academic trends.

Uses
3.4k
Tools
8
Runs
Hosted
Prompts
3
Listed
8mo ago

Get started

Add it to your agent

npx -y @smithery/cli@latest install shawnnygoh/arxiv-scout --client claude

Swap --client claude for cursor, vscode, windsurf or another client. The CLI asks for any required settings.

What it can do

8 tools

arxiv_search_papers8 paramsSearch arXiv for papers using advanced query syntax.
ParameterTypeDescription
querystringThe search keywords or structured arXiv query string (e.g., 'au:vaswani AND ti:attention').
offsetintegerThe number of search results to skip, used for paginating through large result sets.
date_tostringFilter results to papers published on or before this date (format: YYYY-MM-DD).
sort_bystringThe criteria used to order the search results.
date_fromstringFilter results to papers published on or after this date (format: YYYY-MM-DD).
categoriesarrayA list of arXiv category codes (e.g., 'cs.LG') or natural language subject names to filter by.
max_resultsintegerThe maximum number of papers to return in the results list (up to 50).
output_formatstringThe desired format for the tool's response output.
arxiv_get_paper2 paramsFetch complete metadata for a specific arXiv paper.
ParameterTypeDescription
arxiv_idstringThe unique arXiv identifier (e.g., '1706.03762') or the full arXiv.org URL.
output_formatstringChoose between human-readable markdown or structured JSON data.
arxiv_get_papers_batch2 paramsRetrieve metadata for multiple papers in a single call.
ParameterTypeDescription
arxiv_idsarrayA list containing up to 20 arXiv identifiers or full URLs.
output_formatstringThe format of the final response (markdown or JSON).
arxiv_download_and_extract6 paramsExtract full text from a paper's PDF.
ParameterTypeDescription
arxiv_idstringThe arXiv ID of the paper from which to extract text.
end_pageintegerThe 1-indexed page number where text extraction should stop.
start_pageintegerThe 1-indexed page number where text extraction should begin.
max_charactersintegerThe maximum number of characters to return from the PDF to prevent context window overflow.
include_sectionsbooleanWhether to include an automatically detected outline of the paper's sections.
include_referencesbooleanWhether to include the bibliography list extracted from the end of the PDF.
arxiv_get_references3 paramsList the papers that this paper cites.
ParameterTypeDescription
limitintegerThe maximum number of outbound references to retrieve.
arxiv_idstringThe arXiv ID of the paper whose references you wish to explore.
output_formatstringThe desired output format for the reference list.
arxiv_get_citations3 paramsList newer papers that have cited this paper.
ParameterTypeDescription
limitintegerThe maximum number of inbound citations to retrieve.
arxiv_idstringThe arXiv ID of the paper to check for inbound citations.
output_formatstringThe desired output format for the citation list.
arxiv_list_categories2 paramsBrowse the arXiv category taxonomy.
ParameterTypeDescription
groupstringAn optional arXiv group code (e.g., 'cs') or a natural language search term (e.g., 'robotics').
output_formatstringThe desired output format for the category list.
arxiv_cache_stats1 paramView internal cache hit rates and sizes.
ParameterTypeDescription
verbosebooleanOptional flag to show detailed stats.

Ready-made

Prompts

  • summarize_paperGenerate a high-quality structured summary using extracted full text.
  • compare_papersPerform a side-by-side analysis of 2-5 arXiv papers.
  • literature_reviewSynthesize a professional literature review for a research topic.

Readable data

Resources

  • arXiv TaxonomyA full list of arXiv subject category codes and their descriptions.
  • Query Syntax GuideA reference guide for advanced search operators like ti:, au:, and boolean logic.
  • Server InfoAn overview of the server's version, tools, and citations data source.