Add database schema with SQLAlchemy async + Alembic migrations

Set up PostgreSQL database layer with async SQLAlchemy 2.0 and asyncpg driver.
Implements 6 core tables (games, routes, pokemon, route_encounters, nuzlocke_runs,
encounters) with foreign keys, indexes, and an initial Alembic migration.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

.beans/nuzlocke-tracker-l7e3--database-schema-design.md

@@ -1,11 +1,11 @@
 ---
 # nuzlocke-tracker-l7e3
 title: Database Schema Design
-status: todo
+status: completed
 type: task
 priority: normal
 created_at: 2026-02-04T15:46:54Z
-updated_at: 2026-02-04T15:47:23Z
+updated_at: 2026-02-05T12:29:19Z
 parent: nuzlocke-tracker-f5ob
 blocking:
     - nuzlocke-tracker-bkhs
@@ -15,21 +15,27 @@ blocking:
 
 Design and implement the database schema for persistent storage.
 
+## Decisions
+- **Database**: PostgreSQL 16 (already in docker-compose)
+- **ORM**: SQLAlchemy 2.0 (async) with asyncpg driver
+- **Migrations**: Alembic
+- **Async**: Full async — asyncpg + async SQLAlchemy sessions
+
 ## Checklist
-- [ ] Choose database (PostgreSQL, SQLite, MongoDB, etc.)
-- [ ] Set up database connection and ORM/query builder
-- [ ] Design and create tables/collections:
-  - [ ] Games (id, name, generation, region)
-  - [ ] Routes (id, name, game_id, order, encounter_method)
-  - [ ] Pokemon (national_dex, name, types, sprite_url)
-  - [ ] RouteEncounters (route_id, pokemon_id, encounter_rate)
-  - [ ] NuzlockeRuns (id, game_id, settings, started_at, status)
-  - [ ] Encounters (id, run_id, route_id, pokemon_id, nickname, status, caught_at)
-- [ ] Set up migrations system
+- [x] Choose database (PostgreSQL)
+- [x] Set up database connection and ORM/query builder (SQLAlchemy 2.0 async + asyncpg)
+- [x] Design and create tables/collections:
+  - [x] Games (id, name, slug, generation, region, box_art_url, release_year)
+  - [x] Routes (id, name, game_id, order)
+  - [x] Pokemon (id, national_dex, name, types[], sprite_url)
+  - [x] RouteEncounters (id, route_id, pokemon_id, encounter_method, encounter_rate) — unique(route, pokemon, method)
+  - [x] NuzlockeRuns (id, game_id, name, status, rules jsonb, started_at, completed_at)
+  - [x] Encounters (id, run_id, route_id, pokemon_id, nickname, status, catch_level, faint_level, caught_at)
+- [x] Set up migrations system (Alembic)
 - [ ] Add seed data for initial games/routes/Pokémon
-- [ ] Create indexes for common queries
+- [x] Create indexes for common queries (FK indexes on routes, route_encounters, nuzlocke_runs, encounters; status index on nuzlocke_runs)
 
 ## Notes
-- Consider SQLite for simplicity in MVP, can migrate later
 - Use foreign keys for referential integrity
-- Status fields should be enums
+- Status fields should be enums
+- Use PostgreSQL enums or check constraints for status/type fields