diff --git a/AGENTS.md b/AGENTS.md
index f53d45b..7e9c932 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -103,6 +103,8 @@ uv run repub crawl -c repub.toml
 
 - Keep `treefmt.nix`, `flake.nix`, and `pyproject.toml` aligned.
 - Prefer updating the flake-exported package and checks rather than adding ad hoc scripts.
+- Put new SQLite schema objects in numbered files under `repub/sql/` such as `002_*.sql`.
+- For backward-compatible column additions on existing SQLite databases, use Peewee's `playhouse.migrate` helpers instead of raw ad hoc `ALTER TABLE` logic.
 - Do not commit, amend, or stage unrelated files unless explicitly asked.
 - Final verication `nix flake check` must be greenbefore claiming task completeness
 
diff --git a/repub/model.py b/repub/model.py
index 2126d50..5e2cd65 100644
--- a/repub/model.py
+++ b/repub/model.py
@@ -19,6 +19,7 @@ from peewee import (
     SqliteDatabase,
     TextField,
 )
+from playhouse.migrate import SchemaMigrator, migrate
 
 DEFAULT_DB_PATH = Path("republisher.db")
 DATABASE_PRAGMAS = {
@@ -79,45 +80,50 @@ def initialize_database(db_path: str | Path | None = None) -> Path:
     database.init(str(resolved_path), pragmas=DATABASE_PRAGMAS)
     database.connect(reuse_if_open=True)
     try:
-        connection = database.connection()
         for path in schema_paths():
-            connection.executescript(path.read_text(encoding="utf-8"))
-        _ensure_schema(connection)
+            database.connection().executescript(path.read_text(encoding="utf-8"))
+        _run_legacy_migrations()
     finally:
         database.close()
 
     return resolved_path
 
 
-def _ensure_schema(connection: Any) -> None:
-    connection.execute(
-        """
-        CREATE TABLE IF NOT EXISTS app_setting (
-            key TEXT PRIMARY KEY,
-            value TEXT NOT NULL
-        )
-        """
-    )
-
-    job_columns = {
-        row[1] for row in connection.execute("PRAGMA table_info('job')").fetchall()
-    }
+def _run_legacy_migrations() -> None:
+    job_columns = {column.name for column in database.get_columns("job")}
+    operations = []
+    migrator = SchemaMigrator.from_database(database)
     if "convert_images" not in job_columns:
-        connection.execute(
-            """
-            ALTER TABLE job
-            ADD COLUMN convert_images INTEGER NOT NULL DEFAULT 1
-            CHECK (convert_images IN (0, 1))
-            """
+        operations.extend(
+            (
+                migrator.add_column(
+                    "job",
+                    "convert_images",
+                    BooleanField(
+                        default=True,
+                        constraints=[Check("convert_images IN (0, 1)")],
+                    ),
+                ),
+                migrator.add_column_default("job", "convert_images", 1),
+            )
         )
     if "convert_video" not in job_columns:
-        connection.execute(
-            """
-            ALTER TABLE job
-            ADD COLUMN convert_video INTEGER NOT NULL DEFAULT 1
-            CHECK (convert_video IN (0, 1))
-            """
+        operations.extend(
+            (
+                migrator.add_column(
+                    "job",
+                    "convert_video",
+                    BooleanField(
+                        default=True,
+                        constraints=[Check("convert_video IN (0, 1)")],
+                    ),
+                ),
+                migrator.add_column_default("job", "convert_video", 1),
+            )
         )
+    if operations:
+        with database.atomic():
+            migrate(*operations)
 
 
 def source_slug_exists(slug: str) -> bool:
diff --git a/repub/sql/002_app_setting.sql b/repub/sql/002_app_setting.sql
new file mode 100644
index 0000000..060d5ab
--- /dev/null
+++ b/repub/sql/002_app_setting.sql
@@ -0,0 +1,4 @@
+CREATE TABLE IF NOT EXISTS app_setting (
+    key TEXT PRIMARY KEY,
+    value TEXT NOT NULL
+);
diff --git a/tests/test_model.py b/tests/test_model.py
index 1553354..3d0729d 100644
--- a/tests/test_model.py
+++ b/tests/test_model.py
@@ -15,6 +15,7 @@ from repub.model import (
     load_max_concurrent_jobs,
     resolve_database_path,
     save_setting,
+    schema_paths,
 )
 
 
@@ -71,6 +72,39 @@ def test_initialize_database_bootstraps_schema_from_sql_files(tmp_path: Path) ->
         connection.close()
 
 
+def test_initialize_database_applies_newer_sql_files_to_existing_databases(
+    tmp_path: Path,
+) -> None:
+    db_path = tmp_path / "existing.db"
+    connection = sqlite3.connect(db_path)
+    try:
+        connection.executescript(schema_paths()[0].read_text(encoding="utf-8"))
+    finally:
+        connection.close()
+
+    initialize_database(db_path)
+
+    connection = sqlite3.connect(db_path)
+    try:
+        table_names = {
+            row[0]
+            for row in connection.execute(
+                """
+                SELECT name
+                FROM sqlite_master
+                WHERE type = 'table' AND name NOT LIKE 'sqlite_%'
+                """
+            )
+        }
+        assert "app_setting" in table_names
+
+        job_columns = {row[1] for row in connection.execute("PRAGMA table_info('job')")}
+        assert "convert_images" in job_columns
+        assert "convert_video" in job_columns
+    finally:
+        connection.close()
+
+
 def test_initialize_database_configures_sqlite_pragmas(tmp_path: Path) -> None:
     db_path = tmp_path / "pragmas.db"