Add rulebook RAG pipeline and LLM-driven game setup wizard

RAG: switch Postgres to pgvector, chunk and embed the three D&D
rulebooks locally via sentence-transformers, and retrieve relevant
excerpts per DM turn (query = latest player message) to ground the
system prompt. Retrieval runs off the event loop and is capped by a
relevance threshold and a max character budget so it can't blow up
context size or cost.

Game setup wizard: creating a game now opens a short chat where the
DM asks about genre, length, and the player's experience level, then
proposes a name and description via a tool call. The player can edit
both before creating the game. Stateless endpoint — the frontend
carries the conversation, no DB needed since the game doesn't exist
yet.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

backend/app/api/routes_games.py

@@ -8,11 +8,13 @@ from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.auth.users import current_active_user
 from app.db import get_async_session
+from app.llm.game_setup import run_game_setup_turn
 from app.models.character import Character
 from app.models.game import Game, GameParticipant
 from app.models.message import Message
 from app.models.user import User
 from app.schemas.game import GameCreate, GameJoin, GameRead
+from app.schemas.game_setup import GameSetupChatRequest, GameSetupChatResponse
 from app.schemas.message import MessageRead
 from app.ws_tickets import issue_ticket
 
@@ -108,6 +110,15 @@ async def create_game(
     return serialized[0]
 
 
+@router.post("/setup-chat", response_model=GameSetupChatResponse)
+async def game_setup_chat(
+    payload: GameSetupChatRequest,
+    user: User = Depends(current_active_user),
+) -> GameSetupChatResponse:
+    result = await run_game_setup_turn(payload.messages)
+    return GameSetupChatResponse(**result)
+
+
 @router.get("/{game_id}", response_model=GameRead)
 async def get_game(
     game_id: uuid.UUID,