Add claim_type field: distinguish claims vs responses vs replies

Legal documents have 3 types of assertions:
- claim: from appeal documents (כתב ערר)
- response: from original responses (כתב תשובה)
- reply: from supplementary responses (תגובה, השלמת טיעון)

DB: added claim_type column to claims table
Extractor: _infer_claim_type() auto-detects from doc_type + title
Updated existing 113 records: 29 claims, 28 responses, 56 replies

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

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

@@ -114,6 +114,25 @@ async def extract_claims_with_ai(
     return [c for c in claims if "party_role" in c and "claim_text" in c]
 
 
+def _infer_claim_type(doc_type: str, source_name: str) -> str:
+    """Determine claim_type from document type and title.
+
+    - 'claim' = from appeal documents (כתב ערר)
+    - 'response' = from original response documents (כתב תשובה)
+    - 'reply' = from supplementary responses (תגובה, השלמת טיעון)
+    """
+    name_lower = source_name.lower() if source_name else ""
+    if doc_type == "appeal" or "כתב ערר" in name_lower:
+        return "claim"
+    if "כתב תשובה" in name_lower:
+        return "response"
+    if any(kw in name_lower for kw in ["תגובת", "השלמת טיעון", "תגובה"]):
+        return "reply"
+    if doc_type == "response":
+        return "response"
+    return "claim"
+
+
 # ── Regex-based extraction (from existing decisions) ──────────────
 
 PARTY_PATTERNS = [
@@ -229,6 +248,11 @@ async def extract_and_store_claims(
     if not claims:
         return {"status": "no_claims", "total": 0, "source": source_name}
 
+    # Determine claim_type from document type and title
+    claim_type = _infer_claim_type(doc_type, source_name)
+    for c in claims:
+        c["claim_type"] = claim_type
+
     stored = await db.store_claims(case_id, claims, source_document=source_name)
 
     # Summarize by role