Practice area separation: multi-tenant axis across DB, RAG, and UI

Adds two orthogonal columns — practice_area (top-level legal domain:
appeals_committee / national_insurance / labor_law) and appeal_subtype
(building_permit / betterment_levy / compensation_197) — denormalized
into cases, documents, document_chunks, decisions, and style_corpus so
vector searches can filter without JOINs.

Why: the system handles two unrelated sub-domains under the same
appeals committee (1xxx building permits and 8xxx/9xxx betterment/197),
with different rules and writing style. Without a separation axis,
search_similar() and the block-writer's precedent lookup were free to
surface betterment-levy paragraphs while drafting a building-permit
decision — a real risk of cross-domain contamination. The same axis
also lets future domains (national insurance, labor law) coexist
without separate schemas.

Schema (V4 migration in db.py):
- ALTER ... ADD COLUMN IF NOT EXISTS on all five tables + composite
  indexes (practice_area first).
- Idempotent backfill: case_number ~ '^1' → building_permit, '^8' →
  betterment_levy, '^9' → compensation_197; propagated to documents,
  chunks, and decisions via case_id; training-corpus rows (case_id NULL)
  default to appeals_committee.

Code:
- New services/practice_area.py with derive_subtype, validate, and
  is_override + enum constants.
- db.create_case / create_document / store_chunks / create_decision
  inherit practice_area from the parent case (or take an explicit
  override for the case_id=None training corpus).
- db.search_similar and search_similar_paragraphs accept practice_area
  + appeal_subtype filters using the denormalized columns.
- tools/search.py auto-resolves the filter from case_number when given.
- block_writer._build_precedents_context now passes the active case's
  practice_area to search_similar_paragraphs — closes the contamination
  hole for the discussion-block precedent fetch.
- tools/cases.case_create auto-derives subtype from case_number; an
  explicit override that disagrees writes a case_subtype_override entry
  to audit_log so we can spot bad classifications later.
- tools/documents.document_upload_training tags new training material
  with practice_area + subtype end-to-end (corpus, document, chunks).

UI (web/static/index.html + web/app.py):
- New-case wizard gets a practice_area dropdown (others disabled until
  national_insurance / labor_law arrive) and an appeal_subtype dropdown
  with JS auto-fill from the case-number prefix; manual edits stick.
- Case header shows a blue badge with practice_area · subtype.
- CaseCreateRequest plumbs both fields through to cases_tools.case_create.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

mcp-server/src/legal_mcp/tools/documents.py

@@ -105,6 +105,8 @@ async def document_upload_training(
     decision_date: str = "",
     subject_categories: list[str] | None = None,
     title: str = "",
+    practice_area: str = "appeals_committee",
+    appeal_subtype: str = "",
 ) -> str:
     """העלאת החלטה קודמת של דפנה לקורפוס הסגנון (training).
 
@@ -114,10 +116,13 @@ async def document_upload_training(
         decision_date: תאריך ההחלטה (YYYY-MM-DD)
         subject_categories: קטגוריות - אפשר לבחור כמה (בנייה, שימוש חורג, תכנית, היתר, הקלה, חלוקה, תמ"א 38, היטל השבחה, פיצויים 197)
         title: שם המסמך
+        practice_area: תחום משפטי (appeals_committee / national_insurance / labor_law)
+        appeal_subtype: סוג ערר (building_permit / betterment_levy / compensation_197).
+                        ריק = יוסק אוטומטית ממספר ההחלטה
     """
     from datetime import date as date_type
 
-    from legal_mcp.services import extractor, embeddings, chunker
+    from legal_mcp.services import chunker, embeddings, extractor, practice_area as pa
 
     source = Path(file_path)
     if not source.exists():
@@ -126,6 +131,11 @@ async def document_upload_training(
     if not title:
         title = source.stem
 
+    # Resolve subtype: explicit > derived from decision_number > 'unknown'
+    if not appeal_subtype:
+        appeal_subtype = pa.derive_subtype(decision_number, practice_area)
+    pa.validate(practice_area, appeal_subtype)
+
     # Copy to training directory (skip if already there)
     config.TRAINING_DIR.mkdir(parents=True, exist_ok=True)
     dest = config.TRAINING_DIR / source.name
@@ -140,25 +150,29 @@ async def document_upload_training(
     if decision_date:
         d_date = date_type.fromisoformat(decision_date)
 
-    # Add to style corpus
+    # Add to style corpus (tagged by domain so block-writer can filter)
     corpus_id = await db.add_to_style_corpus(
         document_id=None,
         decision_number=decision_number,
         decision_date=d_date,
         subject_categories=subject_categories or [],
         full_text=text,
+        practice_area=practice_area,
+        appeal_subtype=appeal_subtype,
     )
 
     # Chunk and embed for RAG search over training corpus
     chunks = chunker.chunk_document(text)
     if chunks:
-        # Create a document record (no case association)
+        # Create a document record (no case association — tag explicitly)
         doc = await db.create_document(
             case_id=None,
             doc_type="decision",
             title=f"[קורפוס] {title}",
             file_path=str(dest),
             page_count=page_count,
+            practice_area=practice_area,
+            appeal_subtype=appeal_subtype,
         )
         doc_id = UUID(doc["id"])
         await db.update_document(doc_id, extracted_text=text, extraction_status="completed")
@@ -176,7 +190,10 @@ async def document_upload_training(
             }
             for c, emb in zip(chunks, embs)
         ]
-        await db.store_chunks(doc_id, None, chunk_dicts)
+        await db.store_chunks(
+            doc_id, None, chunk_dicts,
+            practice_area=practice_area, appeal_subtype=appeal_subtype,
+        )
 
     return json.dumps({
         "corpus_id": str(corpus_id),