Scanned servers

492 servers from the official registry, last scanned on 2026-10-16. Each axis is scored independently and shown with how much of it could actually be measured — the hatched part of a bar is the share nobody assessed, and it is drawn rather than rounded away.

This set is pinned and only ever grows, so a page published here keeps its URL. It is not a random sample: the servers published first were drawn from the registry's alphabetical head, and everything added since is drawn from the whole registry. 2 of these pages carry a registry note — no longer listed, deprecated by the publisher, or a scan we could not refresh — and each says which.

There is no single number for a server here. The weighted composite is computed and withheld until a hand-audited gold set validates the weights.

Maintenance reads commit history, release cadence, issue response times and contributor spread from the hosting forge. It is scored for 399 of these 492 servers; the rest declare a repository we cannot read, or one on a forge this scanner does not query yet, and each page says which. Its popularity-within-category sub-check is still unassessed for everyone — that needs a distribution across the whole registry, and a category the registry does not record.

Build output (dist/, build/) is not read as a server's own source, because a bundle can inline its dependencies' code, and many npm packages ship nothing else. For 74 of them Code Safety is scored on the repository the server declares instead, and the axis says so; for 15 more no readable repository was declared, so the axis is left unassessed. That is a limit of this scanner, not a fact about those servers.