System Integrity

Every defect this scanner has had was found by someone running a query by hand — a logo field that serialised as an empty object, chain slugs an upstream had renamed, rows dropped from tracking by our own request budget, tokens credited with zero holders while holding liquidity. None of them were an upstream going quiet, so none were caught by uptime monitoring. This page is the answer to that: each past defect is now a rule the system checks on itself, and every violation it finds is published here whether it flatters us or not.

A count is the number of stored rows that break the rule. Zero is a pass. A rule that could not be evaluated shows n/a and holds the overall verdict at “unknown” — a check that did not run has not verified anything, and reporting it as healthy is the failure mode this page exists to prevent.