Returns aggregated statistics — distinct value count and the top-K most
frequent value/count pairs — for one or more dashboard attributes,
optionally restricted to a CEL-filtered subset of dashboards.

Each attributes entry is a CEL expression that produces a comparable
value (string, int, bool, timestamp); compound expressions are allowed.
The CEL environment matches GET /v1/dashboards; see that endpoint's
filter for the available fields.

Each multiValueAttributes entry is a CEL expression that produces a
list<comparable>; the list is flattened across all matching dashboards
and counted as total occurrences, so a single dashboard whose
expression evaluates to a 3-element list contributes 3 increments.
multiValueAttributes count values may therefore exceed
meta.filteredDashboards. For attributes the count is unchanged: the
number of dashboards with that value.

At least one of attributes or multiValueAttributes must be
non-empty; passing neither returns 400.

managedBy is nullable, and a bare managedBy.record.label raises an
evaluation error rather than yielding null, failing the whole request
with 400. Ask with a conditional and choose a sentinel:
managedBy == null ? "null" : managedBy.record.label.

Stat values are returned as JSON strings regardless of the expression's
CEL type:

• string — verbatim.
• int — decimal digits (e.g. "41000234").
• bool — "true" or "false".
• timestamp — RFC 3339 (e.g. "2024-01-01T00:00:00Z").
• a genuine null — the text "null".

The response meta block reports totalDashboards (before the filter)
and filteredDashboards (after) so callers can compute coverage.