Every "Integrity screen: passed" line, every /check result, and every issue links back here. This page names, in plain words, exactly what we check today, what "unknown" means, what we do not check yet, and how to run the same check yourself.
Two things, on every paper Lucent publishes. First, we ask Crossref, live, whether the paper's DOI resolves. Second, we read Crossref's metadata for a retraction-related record. That is the complete list of what ships today. We are not going to describe anything else on this page as a live check, because it would not be true yet.
When you see the words "Integrity screen: passed" next to a paper, it means exactly two things were confirmed at the moment we checked: the DOI resolved, and Crossref's metadata carried no retraction-related field for it. It does not mean the paper's conclusions are correct, it is not a peer-review judgment, and it is not an endorsement of the authors or the journal. We write out what passed in words rather than showing a bare checkmark or a score, because a checkmark alone tells you nothing about what was actually tested.
A check that cannot complete is never shown as a pass. If Crossref has no record for a DOI, or the check cannot be run, we say unknown, plainly, and we say why. The absence of a red flag is not the same thing as a clean result, and we do not let the two look alike. This is the same fail-closed rule whether the check runs inside an issue or on the public checker.
Some things Lucent will never check, and some things we simply have not built yet. Both matter to say out loud:
Lucent's screen calls Crossref's public works API for the paper's DOI. A response confirms the DOI is registered and resolves. We then read three fields in that response: update-to, updated-by, and relation entries typed is-retracted-by or is-retraction-of. If any of those name a retraction, we say so and name the field. If none do, we say the metadata carries no retraction-related field as of the moment we checked, and we timestamp it, because status can change after that moment. Anyone can reproduce this: paste the DOI at doi.org to confirm it resolves, then look at the same Crossref fields, or skip the manual steps and run the same check yourself for free. For the reasoning behind why a retraction matters at all, see what a paper retraction is and how to tell if a study is trustworthy.
None of what follows is live. It is a staged roadmap, described here as future work and nowhere on this site as a current check. Over time we plan to add more mechanical, registry-driven checks, one at a time, with tests, before any copy describes them as running: things like data-availability statements, preprint status, a journal's COPE membership, and PubPeer discussion threads presented strictly as "concerns raised, here is the thread," never as a verdict. Further out, the destination is a published, versioned methodology for journal-level fact sheets, and eventually a reproducible integrity index built entirely on named, third-party-recorded facts. Each stage ships only after its own gates are met, and this page will be updated the day any of it goes live, not before.
Every two days: brand-new papers in plain language, each screened for retractions and a live DOI before you ever see them. Three free every issue.
Check a DOI yourself · What is a paper retraction · How to tell if a study is trustworthy · Latest issue · Privacy