feat(corroboration): halacha matcher + cosine threshold (INV-COR3, X11)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

mcp-server/src/legal_mcp/config.py

@@ -54,6 +54,8 @@ REDIS_URL = os.environ.get("REDIS_URL", "redis://127.0.0.1:6380/0")
 # pinned.
 HALACHA_EXTRACT_MODEL = os.environ.get("HALACHA_EXTRACT_MODEL", "claude-opus-4-8")
 HALACHA_EXTRACT_EFFORT = os.environ.get("HALACHA_EXTRACT_EFFORT", "xhigh")
+HALACHA_CORROBORATION_MATCH_FLOOR = float(os.environ.get("HALACHA_CORROBORATION_MATCH_FLOOR", "0.50"))
+HALACHA_CORROBORATION_MIN_CITES = int(os.environ.get("HALACHA_CORROBORATION_MIN_CITES", "2"))
 
 # Voyage AI
 VOYAGE_API_KEY = os.environ.get("VOYAGE_API_KEY", "")

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

@@ -25,6 +25,14 @@ def _coerce_treatment(raw: dict) -> str:
     return t if t in _VALID_TREATMENT else "mentioned"
 
 
+def accept_match(best: tuple[str, float] | None, floor: float = config.HALACHA_CORROBORATION_MATCH_FLOOR) -> str | None:
+    """Return the halacha_id iff similarity clears the floor (INV-COR3)."""
+    if not best:
+        return None
+    halacha_id, sim = best
+    return halacha_id if sim >= floor else None
+
+
 _TREATMENT_PROMPT = """אתה משפטן בכיר. נתון ציטוט של פסק/החלטה קודמים בתוך החלטה מאוחרת.
 סווג כיצד ההחלטה המאוחרת **מטפלת** בתקדים המצוטט, לפי אחת מהקטגוריות:
 - followed   — אימצה והחילה את ההלכה.

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

@@ -3386,6 +3386,18 @@ async def update_halacha(
     return dict(row) if row else None
 
 
+async def nearest_halacha_for_vector(case_law_id: UUID, vec: list[float]) -> tuple[str, float] | None:
+    """Best-matching halacha of `case_law_id` for a context embedding (cosine)."""
+    pool = await get_pool()
+    row = await pool.fetchrow(
+        "SELECT id::text AS id, 1 - (embedding <=> $2) AS sim "
+        "FROM halachot WHERE case_law_id = $1 AND embedding IS NOT NULL "
+        "ORDER BY embedding <=> $2 LIMIT 1",
+        case_law_id, vec,
+    )
+    return (row["id"], float(row["sim"])) if row else None
+
+
 async def search_precedent_library_semantic(
     query_embedding: list[float],
     practice_area: str = "",

mcp-server/tests/test_corroboration.py

@@ -15,3 +15,12 @@ def test_treatment_polarity():
     assert cor.is_positive("followed") and cor.is_positive("explained")
     assert cor.is_negative("distinguished") and cor.is_negative("overruled")
     assert not cor.is_positive("mentioned") and not cor.is_negative("mentioned")
+
+def test_match_accepts_above_threshold():
+    assert cor.accept_match(("h1", 0.62), floor=0.50) == "h1"
+
+def test_match_rejects_below_threshold():
+    assert cor.accept_match(("h1", 0.41), floor=0.50) is None
+
+def test_match_rejects_empty():
+    assert cor.accept_match(None, floor=0.50) is None