Add `other_count` Rollup to Grouped Aggregate Responses

Grouped aggregate responses (groupby and groupby_time) return the top-N groups ordered by count. Without a rollup field, callers cannot tell whether the top-N is a complete picture of the data or a truncated subset. A model or agent that trusts a limited facet list without knowing the tail size can draw incorrect conclusions (e.g. "these are all the senders" when 40% of records fell into groups beyond the limit). A second aggregation call to compute the tail is extra latency and an unnecessary round trip.

The MCP aggregate tool description SHALL inform models that grouped responses include othercount (the sum of counts for groups/buckets beyond limit) so that truncation can be detected without a second round trip. The tool's text summarizer SHALL include othercount in the model-visible content[] summary when the resource server returns it.

Grouped aggregation responses SHALL enforce a maximum bucket limit and deterministic ordering. Grouped responses SHALL include an othercount field containing the sum of counts for all groups or buckets truncated by limit. othercount SHALL be zero when all groups fit within the limit; it SHALL be positive when truncation occurred. othercount SHALL be present on every grouped response (scalar groupby and calendar groupbytime) and SHALL be absent on ungrouped (scalar) aggregation responses.