Replace all Anthropic API calls with Claude Code session (claude -p)

New module claude_session.py provides query() and query_json() that
run prompts via `claude -p` CLI — uses the claude.ai session, zero API cost.

Converted 6 services:
- claims_extractor.py: extract_claims_with_ai
- brainstorm.py: brainstorm_directions
- block_writer.py: write_block (was streaming+thinking, now simple)
- qa_validator.py: claims_coverage check
- style_analyzer.py: 3 API calls (single pass, multi pass, synthesis)
- learning_loop.py: extract_lessons

Only extractor.py still uses Anthropic API (for PDF OCR with Vision).

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

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

@@ -6,10 +6,8 @@ import json
 import logging
 import re
 
-import anthropic
-
 from legal_mcp import config
-from legal_mcp.services import db
+from legal_mcp.services import db, claude_session
 
 logger = logging.getLogger(__name__)
 
@@ -150,24 +148,16 @@ async def _analyze_single_pass(rows) -> dict:
         decisions_text += f"\n\n--- החלטה {row['decision_number'] or 'ללא מספר'} ---\n"
         decisions_text += row["full_text"]
 
-    client = anthropic.Anthropic(api_key=config.ANTHROPIC_API_KEY)
-    message = client.messages.create(
-        model="claude-opus-4-6",
-        max_tokens=16384,
-        messages=[
-            {
-                "role": "user",
-                "content": ANALYSIS_PROMPT.format(decisions=decisions_text),
-            }
-        ],
+    raw = claude_session.query(
+        ANALYSIS_PROMPT.format(decisions=decisions_text),
+        timeout=claude_session.LONG_TIMEOUT,
     )
 
-    return await _parse_and_store_patterns(message.content[0].text, len(rows))
+    return await _parse_and_store_patterns(raw, len(rows))
 
 
 async def _analyze_multi_pass(rows) -> dict:
     """Analyze each decision individually, then synthesize patterns."""
-    client = anthropic.Anthropic(api_key=config.ANTHROPIC_API_KEY)
     all_patterns = []
 
     # Pass 1: Analyze each decision individually
@@ -175,18 +165,12 @@ async def _analyze_multi_pass(rows) -> dict:
         decision_text = f"--- החלטה {row['decision_number'] or 'ללא מספר'} ---\n"
         decision_text += row["full_text"]
 
-        message = client.messages.create(
-            model="claude-opus-4-6",
-            max_tokens=8192,
-            messages=[
-                {
-                    "role": "user",
-                    "content": SINGLE_DECISION_PROMPT.format(decision=decision_text),
-                }
-            ],
+        raw = claude_session.query(
+            SINGLE_DECISION_PROMPT.format(decision=decision_text),
+            timeout=claude_session.LONG_TIMEOUT,
         )
 
-        patterns = _extract_json(message.content[0].text)
+        patterns = _extract_json(raw)
         if patterns:
             all_patterns.extend(patterns)
 
@@ -194,21 +178,15 @@ async def _analyze_multi_pass(rows) -> dict:
         return {"error": "לא הצלחתי לחלץ דפוסים מההחלטות"}
 
     # Pass 2: Synthesize across all decisions
-    message = client.messages.create(
-        model="claude-opus-4-6",
-        max_tokens=16384,
-        messages=[
-            {
-                "role": "user",
-                "content": SYNTHESIS_PROMPT.format(
-                    num_decisions=len(rows),
-                    patterns=json.dumps(all_patterns, ensure_ascii=False, indent=2),
-                ),
-            }
-        ],
+    raw = claude_session.query(
+        SYNTHESIS_PROMPT.format(
+            num_decisions=len(rows),
+            patterns=json.dumps(all_patterns, ensure_ascii=False, indent=2),
+        ),
+        timeout=claude_session.LONG_TIMEOUT,
     )
 
-    return await _parse_and_store_patterns(message.content[0].text, len(rows))
+    return await _parse_and_store_patterns(raw, len(rows))
 
 
 def _extract_json(response_text: str) -> list | None: