The archive has no public search endpoint, so search here filters the RSS feed, roughly the last 100 papers. That makes it good for tracking what is appearing now and poor for digging up an older paper you half remember. Fetching a known paper by its `2025/123` id works regardless of age.
A reader for the IACR Cryptology ePrint Archive, the preprint server for cryptography research. It works off the archive's RSS feed rather than a search API, which shapes what it can and cannot find.
- Papers searched by title, abstract, author or keyword, with an author filter and a date range — `search_papers`
- One paper's full details by its archive id, in the `2025/123` form — `get_paper`
- The latest published papers straight off the feed — `get_recent_papers`
- Web page and PDF URLs for a paper, and download information for its PDF — `get_paper_url`, `download_paper`
Node.js 18.0.0 or higher. No account, no key, no cookies. Runs as a Node process against the built `dist/index.js`. Two optional variables: `DEBUG` set to 1 for verbose logging, and `MAX_RESULTS` to cap search results, which defaults to 100.
One command — npx -y github:heewon-chung/eprint-mcp-server