MCP connector · Research

Paperfind

Find research papers and open-access links.

  • 12 tools
  • 49 tests passing
  • MIT licence
  • Free and open

InstallStar on GitHubSupport the project

See it work

Example. A real tool and its real input. No live data is shown here.

You ask
“Find open-access papers on malaria vaccines since 2022.”
Your bot calls
search_papers with query, yearFrom and openAccessOnly
It gets back
De-duplicated papers, with a status for each source searched.

Install

git clone https://github.com/si-essence/paperfind-mcp.git && cd paperfind-mcp && npm install && npm run build

Needs Node.js. Then add the built server to your MCP client. Install guide

Facts

Tools
12
Tests
49 passing
Licence
MIT
Source
github.com/si-essence/paperfind-mcp
Listing updated
31 October 2026
Last code push
23 October 2026
Maintained by
Bot Essentials

Tests were run on 27 October 2026 in a clean copy of each connector. You bring your own keys for any service it talks to.

Details and limits

Paperfind is a read-only MCP server for academic paper metadata and links. The package name is papers. You run it over stdio, or over stateless Streamable HTTP on 127.0.0.1.

It searches arXiv, Crossref, OpenAlex, Europe PMC, PubMed, and DOAJ. Unpaywall is enrichment, not a search source. It looks up open-access locations by DOI.

Set PAPERS_CONTACT_EMAIL to an address you control. Unpaywall needs that email. The placeholder contact@example.com is not accepted by Unpaywall. The same address is the User-Agent mailto and is sent to Crossref and PubMed. OPENALEX_API_KEY and NCBI_API_KEY are optional.

It returns PDF URLs as links and refuses to fetch PDF or arXiv source bytes. It does not re-serve PDF bytes. A redirect toward a PDF is refused. Full text is returned only for an open-access Europe PMC record with an open licence, together with that licence.

This process enforces these rates. arXiv is 1 request every 3 seconds, one at a time. Crossref starts at 3 requests per second in the polite pool, then slows when X-Rate-Limit headers are stricter. OpenAlex is 10 requests per second with no key and 30 with a key. Europe PMC is 5 requests per second. PubMed is 3 requests per second, or 10 when NCBI_API_KEY is set. DOAJ is 2 requests per second. Unpaywall is 1 request per second. The inbound window defaults to 60 calls a minute. A cache hit does not spend it.

Semantic Scholar is excluded. Its API licence forbids repackaging, selling, renting, leasing, lending, distributing, or sublicensing the API, some of the data is CC BY-NC, and Ai2 ships an official MCP. Google Scholar is excluded because it has no API for this use and scraping breaks its terms. Sci-Hub and LibGen are excluded because they distribute texts without a licence to do so.

Metadata comes from Crossref. OpenAlex data is CC0. arXiv metadata is CC0. Thank you to arXiv for use of its open access interoperability. Europe PMC records are used through its REST API. This connector uses NCBI E-utilities from the National Library of Medicine (NLM). NLM is not responsible for this project and does not endorse it. DOAJ article metadata is CC0. Unpaywall locations come from the Unpaywall API, now documented by OpenAlex. None of these providers endorse this project.

The connector code is MIT. The data licences above are separate.

Tools

12 tools. Every tool is read-only.

  • search_papers

    Search Crossref, OpenAlex, Europe PMC, PubMed, and DOAJ. arXiv runs only when named.

  • search_source

    Search one named source. An arXiv category is optional. A failure of that source fails the call.

  • get_paper_by_doi

    Resolve one DOI through Crossref and OpenAlex, then add Unpaywall locations when that lookup succeeds.

  • get_paper

    Fetch one paper by arXiv id, PMID, PMCID, or OpenAlex work id.

  • find_open_access

    Open-access locations for a DOI. Unpaywall first, then OpenAlex. Links and licences, never PDF bytes.

  • get_citations

    Works that cite a paper, from OpenAlex and Europe PMC.

  • get_references

    Works a paper references, from OpenAlex, Crossref deposits, and Europe PMC.

  • search_authors

    Search OpenAlex authors. ORCID, affiliations, works count, and h-index when the source provides them.

  • get_author_works

    Works for one OpenAlex author id.

  • get_full_text

    Plain text only for an open-access Europe PMC record with an open licence. Never arXiv full text or PDF bytes.

  • list_sources

    Each source, whether it is searchable, the enforced rate, terms links, and which env names are set. Values are never returned.

  • export_citation

    BibTeX or RIS from supplied metadata, or from a DOI resolved first.