fix(precedents): מראה-מקום never blank — seed at upload + inline Gemini enrichment (root cause)

Root cause: the upload form "מראה המקום" field is stored only as case_number;
citation_formatted was left empty and filled ONLY by the async metadata
extraction, which runs in the local drainer (not the container) because it was
lumped with halacha extraction. Result: the chair saw an empty מראה-מקום during
the window and re-typed it by hand.

Fix (hybrid — Option C):
- Seed: _create_external_record / create_external_case_law store the chair's typed
  citation as citation_formatted at INSERT, so the field is never blank. A
  cited_only→external promotion preserves an existing non-empty value (prior edit).
- Enrich: ingest_precedent runs metadata extraction INLINE in-container
  (gemini_session is direct REST, no local CLI) with force_citation=True, upgrading
  the seed to the canonical derived citation (parties + reporter + date) before the
  chair opens the page. Best-effort: on no key / API failure the seed remains and
  the queued metadata drain stays as the fallback. Halacha extraction is untouched
  (stays local — claude_session).
- apply_to_record / extract_and_apply gain force_citation: re-assemble even when
  citation_formatted is non-empty, writing only when assembly SUCCEEDS (seed
  preserved on abstention). The drainer keeps the default (False) so a chair's
  manual edit in /precedents/[id] is never clobbered.

Ops: GOOGLE_GEMINI_API_KEY added to the legal-ai Coolify app (runtime) — the SoT
value from Infisical nautilus:/external-apis/gemini, validated against the API.

Invariants: G2 (one citation-resolution + one metadata-extraction path, reused —
no parallel logic), INV-ID2/X1§3 (citation_formatted stays a derived field; the
seed is provisional and upgraded), INV-G10 (chair edits preserved). No schema change.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

mcp-server/src/legal_mcp/services/precedent_metadata_extractor.py

@@ -260,6 +260,7 @@ async def apply_to_record(
     case_law_id: UUID | str,
     suggested: dict,
     overwrite_case_number: bool = False,
+    force_citation: bool = False,
 ) -> dict:
     """Merge suggested metadata into the case_law row, filling ONLY empty fields.
 
@@ -274,6 +275,13 @@ async def apply_to_record(
 
     overwrite_case_number: when True, update case_number from case_number_clean
       even if the field already has a value (used for one-time migration enrichment).
+
+    force_citation: when True, (re)assemble citation_formatted even if the field
+      is non-empty — used by the at-upload inline enrichment to UPGRADE the seeded
+      provisional citation (the raw chair input) to the canonical derived form. The
+      write still happens only when the deterministic assembly SUCCEEDS (a missing
+      component → no write → the seed is preserved). The drainer keeps the default
+      (False) so a chair's manual edit in /precedents/[id] is never clobbered.
     """
     if isinstance(case_law_id, str):
         case_law_id = UUID(case_law_id)
@@ -482,7 +490,7 @@ async def apply_to_record(
     # source_type/district/proceeding_type/parties). Only fill when empty so chair
     # edits in /precedents/[id] are preserved; abstains (no write) when a component
     # is missing.
-    if not (record.get("citation_formatted") or "").strip():
+    if force_citation or not (record.get("citation_formatted") or "").strip():
         eff = {**record, **fields_to_update}
         eff_parties = (
             fields_to_update.get("parties") or record.get("parties") or ""
@@ -505,6 +513,7 @@ async def apply_to_record(
 async def extract_and_apply(
     case_law_id: UUID | str,
     overwrite_case_number: bool = False,
+    force_citation: bool = False,
 ) -> dict:
     """Convenience wrapper: extract → merge into row → return summary."""
     suggested = await extract_metadata(case_law_id)
@@ -523,7 +532,11 @@ async def extract_and_apply(
             "status": "extraction_failed" if has_text else "no_metadata",
             "fields": [],
         }
-    result = await apply_to_record(case_law_id, suggested, overwrite_case_number=overwrite_case_number)
+    result = await apply_to_record(
+        case_law_id, suggested,
+        overwrite_case_number=overwrite_case_number,
+        force_citation=force_citation,
+    )
     if result["updated"]:
         await db.recompute_searchable(case_law_id)
     return {