Get the 90-day signal exception strip for a check

Returns the same 90-day history the dashboard's monitor detail page renders as its signal strip, as JSON. Only exception days (down, nodata) are included — an up day emits nothing, so a caller renders every offset from 1 to 90 that is absent from exceptions as up. late is never emitted: this strip distinguishes down vs. up vs. no-data over a calendar day, not the live "late" status a check can be in right now.

nodata is all-or-nothing per check, not per day: it appears when the monitor has no ping history at all — every one of the 90 days comes back nodata in that case, never a mix of nodata and down/absent-meaning-up days. A monitor with any ping history gets ordinary down-or-absent days for the whole window instead.

404 if the check does not exist in the authenticated project (no existence oracle — cross-tenant access is indistinguishable from a non-existent id).

Console stability note. This response is shaped by a console screen: the fields are the ones that screen draws, in the grouping it draws them in. It stays published because reading it is a reasonable thing to automate, and no field will be removed or change meaning without a version. Fields may be ADDED as the screen grows, so treat unknown keys as normal rather than as an error.