add-explore-record-buckets

Why

Explore's over-time chart currently fans out from the console to many per-stream aggregateRecordsByTime calls. That makes first load slow and keeps the chart sparse when the selected corpus spans years because granularity is derived from a fixed client window instead of the populated record extent.

What Changes

• Add an owner-session GET /_ref/explore/records/buckets reference route.
• Reuse the merged timeline substrate's scoped record set and semantic-time expression.
• Return one dense, zero-filled bucket series in a single call, with server-derived populated extent and auto granularity snapped to a calendar ladder targeting roughly 30-60 bars.
• Keep chart honesty: counts are exact reachable-record counts for the requested scope, and empty buckets are explicit zeros.

Capabilities

Modified:

• reference-implementation-architecture

Impact

• Reference implementation server route, operation, contract metadata, and tests.
• No PDPP Core protocol route.
• No grant-scoped /v1 or /mcp behavior change.
• No deployment in this change.