New research

How this is produced

Every Monday a job re-runs this site's standing search queries against the PubMed E-utilities API and Crossref, compares the results against what we already hold, and records anything new. The queries are fixed and public — they live in the repository that builds this site — so the digest measures the literature rather than our attention.

What we do with a new paper

• It appears here, as indexed, with its publication type and a link to the original.
• It joins this site's reference set and becomes available to the pages covering that topic.
• If it changes what a page says, that page is rewritten and its date moves. If it does not, that page keeps its old date — we do not bump dates to manufacture freshness, because a date that always moves carries no information.
• We do not summarise or grade it here. Assessment happens on the page where the claim lives, alongside the rest of the evidence, not in a weekly list.

Why the empty weeks are published

Most weeks, in most mature fields, nothing is published that changes anything. A digest that always has something in it is either drawing on queries so broad they are meaningless, or it is padding. Publishing the quiet weeks is what makes the busy ones informative.

The complete reference set for this site is also available as structured data, and the site map for answer engines is at llms.txt.