perf(search): keyset pagination + fix the format-gate line-ending flip-flop

RECOMMENDATIONS #8: additive AfterSentAtUtc+AfterId cursor — browse windows fetch
via a (SentAtUtc, Id) keyset filter, O(pageSize) at any depth, COUNT skipped
(TotalCount=-1). Adds the Id tie-break to browse ordering (fixes latent
duplicate/skip on equal timestamps). Verified: InMemory continuation test + a
live-Postgres throwaway proving the Guid cursor comparison translates (removed).

Also root-causes the recurring pre-commit/CI format failures: dotnet-format on
Windows defaulted to CRLF while .gitattributes stores LF — the two fought forever.
Adds .editorconfig (end_of_line=lf etc.) so every tool agrees, and normalises the
straggler files. This bug bit three separate commits before being diagnosed.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

src/InboxIntel.Application/DTOs/SearchDtos.cs

@@ -23,4 +23,9 @@ public record SearchRequestDto(
     bool? IsTrashed = null,
     string? GmailLabel = null,   // e.g. "SENT", "DRAFT", "SPAM"
     string? Category = null,     // EmailCategory name, e.g. "Finance"
-    long? MinSizeBytes = null);
+    long? MinSizeBytes = null,
+    // Keyset cursor for the date-ordered browse path (RECOMMENDATIONS #8): pass the last
+    // row's SentAtUtc+Id to fetch the next window without OFFSET (O(pageSize), not O(page)).
+    // When set, TotalCount is not recomputed (-1). Additive; offset paging still works.
+    DateTimeOffset? AfterSentAtUtc = null,
+    Guid? AfterId = null);