feat: add animal registry schema and models

Add database tables for animal tracking:
- animal_registry: main snapshot table with all animal attributes
- live_animals_by_location: denormalized view for fast roster queries
- animal_aliases: merge tracking for when animals are discovered to be same

Includes Pydantic models and comprehensive tests for all constraints.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

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

migrations/0003-animal-registry-schema.sql

@@ -0,0 +1,58 @@
+-- ABOUTME: Creates animal tracking tables for the registry system.
+-- ABOUTME: Includes animal_registry, live_animals_by_location, and animal_aliases.
+
+-- Main snapshot table for all animals (current state)
+CREATE TABLE animal_registry (
+    animal_id TEXT PRIMARY KEY CHECK(length(animal_id) = 26),
+    species_code TEXT NOT NULL REFERENCES species(code),
+    identified INTEGER NOT NULL DEFAULT 0 CHECK(identified IN (0, 1)),
+    nickname TEXT,
+    sex TEXT NOT NULL CHECK(sex IN ('male', 'female', 'unknown')),
+    repro_status TEXT NOT NULL CHECK(repro_status IN ('intact', 'wether', 'spayed', 'unknown')),
+    life_stage TEXT NOT NULL CHECK(life_stage IN ('hatchling', 'juvenile', 'subadult', 'adult')),
+    status TEXT NOT NULL CHECK(status IN ('alive', 'dead', 'harvested', 'sold', 'merged_into')),
+    location_id TEXT NOT NULL REFERENCES locations(id),
+    origin TEXT NOT NULL CHECK(origin IN ('hatched', 'purchased', 'rescued', 'unknown')),
+    born_or_hatched_at INTEGER,
+    acquired_at INTEGER,
+    first_seen_utc INTEGER NOT NULL,
+    last_event_utc INTEGER NOT NULL
+);
+
+-- Unique nickname only for active animals (allows nulls, allows reuse for dead/merged)
+CREATE UNIQUE INDEX idx_ar_nickname_active
+    ON animal_registry(nickname)
+    WHERE nickname IS NOT NULL
+    AND status NOT IN ('dead', 'harvested', 'sold', 'merged_into');
+
+CREATE INDEX idx_ar_location ON animal_registry(location_id);
+CREATE INDEX idx_ar_filter ON animal_registry(species_code, sex, life_stage, identified);
+CREATE INDEX idx_ar_status ON animal_registry(status);
+CREATE INDEX idx_ar_last_event ON animal_registry(last_event_utc);
+
+-- Denormalized view for fast roster queries (only alive animals)
+CREATE TABLE live_animals_by_location (
+    animal_id TEXT PRIMARY KEY CHECK(length(animal_id) = 26),
+    location_id TEXT NOT NULL REFERENCES locations(id),
+    species_code TEXT NOT NULL REFERENCES species(code),
+    identified INTEGER NOT NULL DEFAULT 0 CHECK(identified IN (0, 1)),
+    nickname TEXT,
+    sex TEXT NOT NULL CHECK(sex IN ('male', 'female', 'unknown')),
+    repro_status TEXT NOT NULL CHECK(repro_status IN ('intact', 'wether', 'spayed', 'unknown')),
+    life_stage TEXT NOT NULL CHECK(life_stage IN ('hatchling', 'juvenile', 'subadult', 'adult')),
+    first_seen_utc INTEGER NOT NULL,
+    last_move_utc INTEGER,
+    tags TEXT NOT NULL DEFAULT '[]' CHECK(json_valid(tags))
+);
+
+CREATE INDEX idx_labl_location ON live_animals_by_location(location_id);
+CREATE INDEX idx_labl_filter ON live_animals_by_location(location_id, species_code, sex, life_stage, identified);
+
+-- Tracks when animals are discovered to be the same individual (merge)
+CREATE TABLE animal_aliases (
+    alias_animal_id TEXT PRIMARY KEY CHECK(length(alias_animal_id) = 26),
+    survivor_animal_id TEXT NOT NULL CHECK(length(survivor_animal_id) = 26),
+    merged_at_utc INTEGER NOT NULL
+);
+
+CREATE INDEX idx_aa_survivor ON animal_aliases(survivor_animal_id);