fix: align production defaults and close review parity gaps

Ship greenfield min_rr=2.0 and conf=0, read-only Structural S/R, indicator
cache invalidation, and UI/gate language that treats GTL as screening not exit.
Align strategy_rank missing-vol fallback live vs backtest, single-source
PRIMARY_TARGET_MIN_RR, expand prod parity tests, and drop dead FE clients.

tests/unit/test_rr_scanner_integration.py

@@ -1,7 +1,8 @@
-"""Integration tests for R:R scanner full flow with quality-based target selection.
+"""Integration tests for R:R scanner full flow with probability-based primary.
 
-Verifies the complete scan_ticker pipeline: quality-based S/R level selection,
-correct TradeSetup field population, and database persistence.
+Verifies scan_ticker → enhance_trade_setup: headline target is the primary
+selected by probability floors (not the pre-enhance quality candidate loop),
+TradeSetup fields, and persistence.
 
 **Validates: Requirements 2.1, 2.2, 2.3, 2.4, 3.4**
 """
@@ -63,35 +64,52 @@ def _make_ohlcv_bars(
 
 
 # ===========================================================================
-# 8.1 Integration test: full scan_ticker flow with quality-based selection,
+# 8.1 Integration test: full scan_ticker flow with probability primary,
 #     correct TradeSetup fields, and database persistence
 # ===========================================================================
 
+def _assert_headline_is_probability_primary(setup: TradeSetup) -> None:
+    """Headline target/rr must match the starred primary from _select_primary_target."""
+    targets = setup.targets or []
+    assert targets, "expected generated targets after enhance"
+    primaries = [t for t in targets if t.get("is_primary")]
+    assert len(primaries) == 1, "exactly one primary target expected"
+    primary = primaries[0]
+    assert setup.target == pytest.approx(float(primary["price"]), abs=0.01)
+    assert setup.rr_ratio == pytest.approx(float(primary["rr_ratio"]), abs=0.01)
+    worthwhile = [
+        t for t in targets
+        if float(t.get("rr_ratio", 0.0)) >= 1.5 and float(t.get("probability", 0.0)) >= 20.0
+    ]
+    pool = worthwhile or targets
+    best = max(pool, key=lambda t: (float(t["probability"]), float(t["rr_ratio"])))
+    assert primary["price"] == pytest.approx(float(best["price"]), abs=0.01)
+
+
 @pytest.mark.asyncio
-async def test_scan_ticker_full_flow_quality_selection_and_persistence(
+async def test_scan_ticker_full_flow_probability_primary_and_persistence(
     scan_session: AsyncSession,
 ):
-    """Integration test for the complete scan_ticker pipeline.
+    """Integration test for the complete scan_ticker → enhance pipeline.
 
     Scenario:
     - Entry ≈ 100, ATR ≈ 2.0, risk ≈ 3.0 (atr_multiplier=1.5)
     - 3 resistance levels above (long candidates):
-        A: price=105, strength=90 (strong, near)  → highest quality
+        A: price=105, strength=90 (strong, near)  → typically highest reach-prob
         B: price=115, strength=40 (medium, mid)
-        C: price=135, strength=5  (weak, far)
+        C: price=135, strength=5  (weak, far / lottery)
     - 3 support levels below (short candidates):
-        D: price=95,  strength=85 (strong, near)  → highest quality
+        D: price=95,  strength=85 (strong, near)
         E: price=85,  strength=35 (medium, mid)
         F: price=65,  strength=8  (weak, far)
     - CompositeScore: 72.5
 
     Verifies:
     1. Both long and short setups are produced
-    2. Long target = Level A (highest quality, not most distant)
-    3. Short target = Level D (highest quality, not most distant)
-    4. All TradeSetup fields are correct and rounded to 4 decimals
-    5. rr_ratio is the actual R:R of the selected level
-    6. Old setups are deleted, new ones persisted
+    2. Headline is the probability-based primary (not a distant lottery)
+    3. Near/strong levels win over far/weak when they clear floors
+    4. rr_ratio matches the selected primary's R:R
+    5. Old setups are deleted, new ones persisted
     """
     # -- Setup: create ticker --
     ticker = Ticker(symbol="INTEG")
@@ -172,16 +190,14 @@ async def test_scan_ticker_full_flow_quality_selection_and_persistence(
     long_setup = long_setups[0]
     short_setup = short_setups[0]
 
-    # -- Assert: long target is Level A (highest quality, not most distant) --
-    # Level A: price=105 (strong, near) should beat Level C: price=135 (weak, far)
+    # -- Assert: headline is probability primary; near/strong beats far lottery --
+    _assert_headline_is_probability_primary(long_setup)
+    _assert_headline_is_probability_primary(short_setup)
     assert long_setup.target == pytest.approx(105.0, abs=0.01), (
-        f"Long target should be 105.0 (highest quality), got {long_setup.target}"
+        f"Long primary should be 105.0 (near, high reach-prob), got {long_setup.target}"
     )
-
-    # -- Assert: short target is Level D (highest quality, not most distant) --
-    # Level D: price=95 (strong, near) should beat Level F: price=65 (weak, far)
     assert short_setup.target == pytest.approx(95.0, abs=0.01), (
-        f"Short target should be 95.0 (highest quality), got {short_setup.target}"
+        f"Short primary should be 95.0 (near, high reach-prob), got {short_setup.target}"
     )
 
     # -- Assert: entry_price is the last close (≈ 100) --