feat: add hand detail API and enrich hand summary fields

- HandSummary: add hole_cards, starting_stacks, ending_stacks, pot_contributions
- Engine: capture all players' hole cards (not just showdown), pre/post hand stacks, per-level pot contributions
- Server: new GET /game/<game_id>/hands/<hand_number> route
- Service: add get_hand_state() method
- Tests: add ServerTests for new endpoint, update existing tests
- Existing GET /game/<game_id> auto-inherits new fields via shared to_dict()

texas_holdem/models.py

@@ -178,6 +178,10 @@ class HandSummary:
     # on the summary (rather than only on the game) guarantees historical
     # hands remain self-describing even after the blinds are raised later.
     blinds: BlindLevel | None = None
+    hole_cards: dict[str, list[Card]] = field(default_factory=dict)
+    starting_stacks: dict[str, int] = field(default_factory=dict)
+    ending_stacks: dict[str, int] = field(default_factory=dict)
+    pot_contributions: list[dict[str, Any]] = field(default_factory=list)
     showdown_hands: dict[str, list[Card]] = field(default_factory=dict)
     started_at: float = field(default_factory=time)
     finished_at: float = field(default_factory=time)
@@ -191,6 +195,16 @@ class HandSummary:
             "board": [str(card) for card in self.board],
             "actions": [record.to_dict() for record in self.actions],
             "awards": [award.to_dict() for award in self.awards],
+            "hole_cards": {
+                player_id: [str(card) for card in cards]
+                for player_id, cards in self.hole_cards.items()
+            },
+            "starting_stacks": dict(self.starting_stacks),
+            "ending_stacks": dict(self.ending_stacks),
+            "pot_contributions": [
+                self._pot_contribution_to_dict(contribution)
+                for contribution in self.pot_contributions
+            ],
             # ``showdown_hands`` is only populated when more than one player
             # remained eligible for a pot; empty dict means the hand ended
             # without a showdown (e.g. everyone folded but the winner).
@@ -201,3 +215,26 @@ class HandSummary:
             "started_at": self.started_at,
             "finished_at": self.finished_at,
         }
+
+    @staticmethod
+    def _pot_contribution_to_dict(contribution: dict[str, Any]) -> dict[str, object]:
+        hand_value = contribution.get("hand_value")
+        if isinstance(hand_value, HandValue):
+            hand_value = hand_value.to_dict()
+        elif isinstance(hand_value, dict):
+            hand_value = dict(hand_value)
+
+        raw_contributors = contribution.get("contributors") or {}
+        contributors = {
+            str(player_id): int(amount)
+            for player_id, amount in dict(raw_contributors).items()
+        }
+        return {
+            "amount": int(contribution.get("amount") or 0),
+            "contributors": contributors,
+            "winners": [
+                str(player_id)
+                for player_id in contribution.get("winners", [])
+            ],
+            "hand_value": hand_value,
+        }