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/learning_loop.py

@@ -12,23 +12,12 @@ from __future__ import annotations
 import logging
 from uuid import UUID
 
-import anthropic
-
 from legal_mcp import config
 from legal_mcp.config import parse_llm_json
-from legal_mcp.services import db
+from legal_mcp.services import db, claude_session
 
 logger = logging.getLogger(__name__)
 
-_anthropic_client: anthropic.Anthropic | None = None
-
-
-def _get_anthropic() -> anthropic.Anthropic:
-    global _anthropic_client
-    if _anthropic_client is None:
-        _anthropic_client = anthropic.Anthropic(api_key=config.ANTHROPIC_API_KEY)
-    return _anthropic_client
-
 
 def compute_diff_stats(draft_text: str, final_text: str) -> dict:
     """חישוב סטטיסטיקות השוואה בין טיוטה לסופית."""
@@ -93,25 +82,15 @@ async def analyze_changes(draft_text: str, final_text: str) -> dict:
     draft_sample = draft_text[:max_chars]
     final_sample = final_text[:max_chars]
 
-    client = _get_anthropic()
-    message = client.messages.create(
-        model="claude-sonnet-4-20250514",
-        max_tokens=4096,
-        messages=[{
-            "role": "user",
-            "content": f"""{LESSONS_PROMPT}
+    prompt = f"""{LESSONS_PROMPT}
 
 --- טיוטה ---
 {draft_sample}
 
 --- גרסה סופית ---
 {final_sample}
-""",
-        }],
-    )
-
-    raw = message.content[0].text.strip()
-    result = parse_llm_json(raw)
+"""
+    result = claude_session.query_json(prompt, timeout=120)
     if result is None:
         logger.warning("Failed to parse lessons response")
         return {"changes": [], "new_expressions": [], "overall_assessment": raw[:200]}