fix(documents): keep an incomplete extraction on the document

A long document read in parts can end with a part cut at the provider's
output limit, a part never read, or more facts than one review takes.
That was told once, as a notice when the run finished; a review opened
later looked complete.

InboundDocument.extractionIncomplete (migration 0483) is written with
the facts by both stagings, the extract route or run and lab
auto-staging. The document detail publishes it and the review shows the
same note whenever it is opened. The disaster-recovery backup carries
it; an older file restores it as false.

Refs #1257

src/lib/validations/inbound-documents.ts

@@ -398,12 +398,12 @@ export interface InboundDocumentDetailDto extends InboundDocumentDto {
   /** When the background summary was generated (ISO 8601), or null. */
   summaryGeneratedAt: string | null;
   /**
-   * What became of the summary. A null `summary` is ambiguous on its own —
-   * never attempted, mid-flight, withheld by the safety screen and could-not-
-   * produce all look identical — so the view reads this instead of guessing
-   * "still generating".
+   * What became of the summary: a null `summary` alone cannot tell never
+   * attempted, mid-flight, withheld and could-not-produce apart.
    */
   summaryState: DocumentSummaryStateValue;
+  /** The staged facts may leave part of the document out (last extraction). */
+  extractionIncomplete: boolean;
 }
 
 // ─── Edit (correction before approval) ─────────────────────────────────────