diff --git a/MALKHUT/malkhut/training/cma_trainer.py b/MALKHUT/malkhut/training/cma_trainer.py
index 671dcd2..205e829 100644
--- a/MALKHUT/malkhut/training/cma_trainer.py
+++ b/MALKHUT/malkhut/training/cma_trainer.py
@@ -1046,6 +1046,7 @@ class PolicyEvaluator:
             # TIE-IN: Record strategy × regime performance
             if record_to_matrix and matrix is not None:
                 tags = scenario.tags
+                venue_tag = scenario.venue
                 for tag in tags:
                     matrix.record(
                         strategy_id=params.version,
@@ -1054,6 +1055,7 @@ class PolicyEvaluator:
                         pnl_bps=result.pnl_bps,
                         drawdown_bps=result.max_drawdown_bps,
                         adverse_fill_ratio=result.adverse_fill_count / max(result.order_count, 1),
+                        venue=venue_tag,
                     )
         score = self._robust_score(results, params)
         return score, results
diff --git a/MALKHUT/malkhut/training/selector.py b/MALKHUT/malkhut/training/selector.py
index e386e0d..4c4024f 100644
--- a/MALKHUT/malkhut/training/selector.py
+++ b/MALKHUT/malkhut/training/selector.py
@@ -188,8 +188,9 @@ class PerformanceMatrix:
         pnl_bps: float = 0.0,
         drawdown_bps: float = 0.0,
         adverse_fill_ratio: float = 0.0,
+        venue: str = "bingx",
     ) -> None:
-        """Record a strategy's performance in a regime."""
+        """Record a strategy's performance in a regime (optionally venue-tagged)."""
         key = (regime, strategy_id)
         existing = self._scores.get(key)