feat: auto-convert camelCase property names to snake_case column names

.claude/plans/camel-to-snake-columns/001-core-camel-to-snake.md

@@ -0,0 +1,33 @@
+# Task 001: Add camelToSnakeCase and Update EntityMetadataFactory
+
+**Status**: completed
+**Depends on**: none
+**Retry count**: 0
+
+## Description
+Add a private `camelToSnakeCase(string $name): string` method to `EntityMetadataFactory` and update the `parse()` method to auto-convert property names to snake_case when no explicit `name:` is provided on the `#[Column]` attribute.
+
+## Context
+- Key file: `packages/database/src/Entity/EntityMetadataFactory.php`
+- Line 68 is the critical change: `$columnName = $columnAttr->name ?? $propertyName;` becomes `$columnName = $columnAttr->name ?? $this->camelToSnakeCase($propertyName);`
+- Test file: `packages/database/tests/Entity/EntityMetadataFactoryTest.php`
+- The existing test "uses Column attribute name when specified" (line 172) tests explicit `name:` override — it should still pass
+- Single-word properties like `$id`, `$name`, `$email` should be unaffected (already lowercase)
+
+## Requirements (Test Descriptions)
+- [ ] `it converts camelCase property names to snake_case column names automatically`
+- [ ] `it preserves explicit Column name override when specified`
+- [ ] `it handles single-word property names without change`
+- [ ] `it handles consecutive uppercase letters correctly (userID becomes user_id)`
+- [ ] `it handles leading uppercase sequences correctly (HTMLParser becomes html_parser)`
+- [ ] `it updates the existing override test to use a genuinely custom name`
+
+## Acceptance Criteria
+- All requirements have passing tests
+- `camelToSnakeCase` handles: `postId` -> `post_id`, `createdAt` -> `created_at`, `id` -> `id`, `HTMLParser` -> `html_parser`, `userID` -> `user_id`, `isActive` -> `is_active`
+- Explicit `#[Column(name: 'custom')]` still takes priority over auto-conversion
+- Existing EntityMetadataFactoryTest tests still pass (some may need column name assertion updates)
+- The existing "uses Column attribute name when specified" test (line 172) MUST be updated to use a genuinely custom name that differs from what auto-conversion would produce (e.g., `#[Column(name: 'author')]` on `$userId` instead of `#[Column(name: 'user_id')]`), otherwise the test becomes a no-op that passes whether or not the override works
+
+## Implementation Notes
+(Left blank - filled in by programmer during implementation)

.claude/plans/camel-to-snake-columns/002-database-unit-tests.md

@@ -0,0 +1,33 @@
+# Task 002: Update Database Package Unit Tests
+
+**Status**: completed
+**Depends on**: 001
+**Retry count**: 0
+
+## Description
+Update unit tests in the database package that assert column names from parsed metadata. The core change causes camelCase property names to produce snake_case column names, so assertions checking column names need updating.
+
+## Context
+- `packages/database/tests/Entity/EntityMetadataFactoryTest.php` — Some test entities use camelCase properties without explicit names. The "extracts #[Column] attributes" test uses `$title`, `$content` (single-word, no change). The "extracts foreign key reference" test uses `$userId` without explicit name — its column will now be `user_id` instead of `userId`. But this test only checks `references`/`onDelete`/`onUpdate`, not the column name, so it should still pass.
+- `packages/database/tests/Entity/SchemaBuilderTest.php` — The "builds ForeignKey objects from column references" test (line 138) expects FK name `fk_posts_userId` and columns `['userId']`. After the change, `$userId` (no explicit name) auto-converts to `user_id`, so expect `fk_posts_user_id` and `['user_id']`. The "preserves foreign key references" test (line 119) doesn't check column name directly — only references/onDelete/onUpdate — so it should pass.
+- `packages/database/tests/Entity/EntityMetadataTest.php` — Manually constructs PropertyMetadata/ColumnMetadata, doesn't use the factory. No changes needed.
+- `packages/database/tests/Entity/EntityHydratorTest.php` — Manually constructs metadata with snake_case column names. No changes needed.
+- `packages/database/tests/Attributes/ColumnAttributeTest.php` — Tests the Column attribute class directly, not the factory. No changes needed.
+- `packages/database/tests/Entity/EntityDiscoveryTest.php` — Uses `$id` only. No changes.
+- `packages/database/tests/Schema/SchemaRegistryTest.php` — Uses single-word properties only. No changes.
+
+## Requirements (Test Descriptions)
+- [ ] `it generates FK name using snake_case column name (fk_posts_user_id)`
+- [ ] `it uses snake_case column names in FK column arrays`
+- [ ] `it preserves column name assertions for single-word properties`
+- [ ] `it passes all existing EntityMetadataFactory tests after column name conversion`
+
+## Acceptance Criteria
+- All tests in `packages/database/tests/Entity/` pass
+- All tests in `packages/database/tests/Schema/` pass
+- All tests in `packages/database/tests/Attributes/` pass
+- FK name in SchemaBuilderTest updated from `fk_posts_userId` to `fk_posts_user_id`
+- FK columns in SchemaBuilderTest updated from `['userId']` to `['user_id']`
+
+## Implementation Notes
+(Left blank - filled in by programmer during implementation)

.claude/plans/camel-to-snake-columns/003-database-feature-tests.md

@@ -0,0 +1,42 @@
+# Task 003: Update Database Package Feature Tests
+
+**Status**: completed
+**Depends on**: 001
+**Retry count**: 0
+
+## Description
+Update feature tests in the database package that use inline entity classes with camelCase properties. After the core change, these properties produce snake_case column names, breaking assertions that reference the old camelCase column names.
+
+## Context
+- `packages/database/tests/Feature/EntityToMigrationWorkflowTest.php`:
+  - `WorkflowUser` has `$isActive` with bare `#[Column]` — column becomes `is_active` instead of `isActive`
+  - `WorkflowPost` has `$authorId` with bare `#[Column]` — column becomes `author_id` instead of `authorId`
+  - Line 230: `->toContain('isActive')` must become `->toContain('is_active')`
+  - Line 237: `->toContain('authorId')` must become `->toContain('author_id')`
+  - Line 267: `new SchemaColumn(name: 'isActive', type: 'BOOLEAN')` must become `name: 'is_active'`
+  - Line 289: `->toContain('isActive')` must become `->toContain('is_active')`
+  - IMPORTANT: Search the entire file for ALL `SchemaColumn` and `SchemaTable` constructions that represent entity-derived schema. Any manually constructed `SchemaColumn` with a camelCase `name:` that represents what the entity would produce must also be updated (e.g., `name: 'isActive'` -> `name: 'is_active'`, `name: 'authorId'` -> `name: 'author_id'`). The "detects and generates migrations" test constructs expected schema objects that must match the new auto-converted names.
+
+- `packages/database/tests/Feature/RepositoryCrudTest.php`:
+  - `CrudProduct` has `$isAvailable` with bare `#[Column]` — column becomes `is_available` instead of `isAvailable`
+  - Line 107: Mock storage uses `'isAvailable'` key — must become `'is_available'`
+  - Lines 203-205: Mock storage arrays use `'isAvailable'` key — must become `'is_available'`
+  - Line 228: `str_contains($sql, 'isAvailable = ?')` must become `str_contains($sql, 'is_available = ?')`
+  - Line 370: Mock storage uses `'isAvailable'` key — must become `'is_available'`
+
+## Requirements (Test Descriptions)
+- [ ] `it uses snake_case column names in entity-to-migration workflow assertions`
+- [ ] `it uses snake_case column names in SchemaColumn construction for diff tests`
+- [ ] `it uses snake_case column names in mock storage arrays for CRUD tests`
+- [ ] `it uses snake_case column names in SQL string matching for findBy tests`
+- [ ] `it passes all feature tests after column name updates`
+
+## Acceptance Criteria
+- All tests in `packages/database/tests/Feature/` pass
+- All `isActive` column references updated to `is_active`
+- All `authorId` column references updated to `author_id`
+- All `isAvailable` column references updated to `is_available`
+- Mock connection storage arrays use snake_case keys for column names
+
+## Implementation Notes
+(Left blank - filled in by programmer during implementation)

.claude/plans/camel-to-snake-columns/004-repository-tests.md

@@ -0,0 +1,40 @@
+# Task 004: Update Repository Tests
+
+**Status**: completed
+**Depends on**: 001
+**Retry count**: 0
+
+## Description
+Update Repository unit tests that use entities with camelCase properties. After the core change, the Repository generates SQL using snake_case column names, so mock connections and assertions need updating.
+
+## Context
+- `packages/database/tests/Repository/RepositoryTest.php`:
+  - `RepositoryTestUser` has `$isActive` with bare `#[Column]` — column becomes `is_active` instead of `isActive`
+  - `$email` with `#[Column('email_address')]` — explicitly named, no change needed
+  - There are **23 occurrences** of `isActive` in this file that need review. Affected areas:
+    - Line 188: Comment `'isActive' maps to 'isActive' column` — update comment to reflect snake_case
+    - Lines 194, 233, 282: Mock connection storage arrays with `'isActive'` key — change to `'is_active'`
+    - Lines 310, 311, 327, 328, 343: Mock query return arrays with `'isActive'` key — change to `'is_active'`
+    - Line 334: `findBy(['isActive' => true])` — this uses the PHP property name as the criteria key, which Repository::findBy maps to column name via `$propertyToColumn`. The call stays as `findBy(['isActive' => true])` (property name), but the generated SQL will now contain `is_active = ?` instead of `isActive = ?`, so mock SQL matching must update accordingly.
+    - Lines 412, 452, 527, 626, 665, 724, 819: Various mock data arrays with `'isActive'` key — change to `'is_active'`
+    - Line 577: `->not->toContain('isActive')` — this checks SQL SET clause; update to `'is_active'`
+  - NOTE: Lines that reference `$user->isActive` (PHP property access) do NOT change — only column name string keys in mock data arrays and SQL assertions change
+
+- `packages/database/tests/Repository/RepositoryLifecycleEventTest.php`:
+  - `LifecycleTestItem` only has `$id` and `$name` (single-word properties) — no changes needed
+  - Verify test still passes after the core change
+
+## Requirements (Test Descriptions)
+- [ ] `it uses snake_case column names in Repository SQL generation assertions`
+- [ ] `it uses snake_case column names in mock connection storage arrays`
+- [ ] `it preserves explicit Column name override (email_address) in Repository tests`
+- [ ] `it passes all Repository tests after column name updates`
+
+## Acceptance Criteria
+- All tests in `packages/database/tests/Repository/` pass
+- All `isActive` column references in mock code updated to `is_active`
+- Explicit `email_address` mapping unchanged
+- RepositoryLifecycleEventTest passes without changes
+
+## Implementation Notes
+(Left blank - filled in by programmer during implementation)

.claude/plans/camel-to-snake-columns/005-production-entity-cleanup.md

@@ -0,0 +1,114 @@
+# Task 005: Remove Redundant Names from Production Entities and Update Entity Tests
+
+**Status**: completed
+**Depends on**: 001
+**Retry count**: 0
+
+## Description
+Remove all explicit `#[Column(name: '...')]` parameters that are now redundant because the auto-conversion produces the same snake_case name. Update entity attribute tests in admin-auth, authentication-token, and media that check Column attribute `name` via reflection.
+
+## Context
+All explicit names in production entities match what the auto-conversion would produce. They are now unnecessary boilerplate.
+
+**Entities to clean up:**
+
+- `packages/admin-auth/src/Entity/AdminUser.php`:
+  - `#[Column('remember_token')]` on `$rememberToken` → `#[Column]`
+  - `#[Column('is_active', default: '1')]` on `$isActive` → `#[Column(default: '1')]`
+  - `#[Column('created_at')]` on `$createdAt` → `#[Column]`
+  - `#[Column('updated_at')]` on `$updatedAt` → `#[Column]`
+
+- `packages/admin-auth/src/Entity/Role.php`:
+  - `#[Column('is_super_admin', default: '0')]` on `$isSuperAdmin` → `#[Column(default: '0')]`
+  - `#[Column('created_at')]` on `$createdAt` → `#[Column]`
+  - `#[Column('updated_at')]` on `$updatedAt` → `#[Column]`
+
+- `packages/admin-auth/src/Entity/Permission.php`:
+  - `#[Column('created_at')]` on `$createdAt` → `#[Column]`
+
+- `packages/admin-auth/src/Entity/RolePermission.php`:
+  - `#[Column('role_id', references: 'roles.id', onDelete: 'CASCADE')]` on `$roleId` → `#[Column(references: 'roles.id', onDelete: 'CASCADE')]`
+  - `#[Column('permission_id', references: 'permissions.id', onDelete: 'CASCADE')]` on `$permissionId` → `#[Column(references: 'permissions.id', onDelete: 'CASCADE')]`
+
+- `packages/media/src/Entity/Media.php`:
+  - `#[Column('original_filename', length: 255)]` on `$originalFilename` → `#[Column(length: 255)]`
+  - `#[Column('mime_type', length: 100)]` on `$mimeType` → `#[Column(length: 100)]`
+  - `#[Column('created_at')]` on `$createdAt` → `#[Column]`
+  - `#[Column('updated_at')]` on `$updatedAt` → `#[Column]`
+
+- `packages/media/src/Entity/MediaAttachment.php`:
+  - `#[Column('media_id')]` on `$mediaId` → `#[Column]`
+  - `#[Column('attachable_type', length: 255)]` on `$attachableType` → `#[Column(length: 255)]`
+  - `#[Column('attachable_id', length: 255)]` on `$attachableId` → `#[Column(length: 255)]`
+
+- `packages/webhook/src/Entity/WebhookAttempt.php`:
+  - `#[Column(name: 'status_code')]` on `$statusCode` → `#[Column]`
+  - `#[Column(name: 'response_body', type: 'TEXT')]` on `$responseBody` → `#[Column(type: 'TEXT')]`
+  - `#[Column(name: 'error_message', type: 'TEXT')]` on `$errorMessage` → `#[Column(type: 'TEXT')]`
+  - `#[Column(name: 'attempted_at')]` on `$attemptedAt` → `#[Column]`
+  - `#[Column(name: 'webhook_url')]` on `$webhookUrl` → `#[Column]`
+  - `#[Column(name: 'attempt_number')]` on `$attemptNumber` → `#[Column]`
+
+- `packages/authentication-token/src/Entity/PersonalAccessToken.php`:
+  - `#[Column('tokenable_type')]` on `$tokenableType` → `#[Column]`
+  - `#[Column('tokenable_id')]` on `$tokenableId` → `#[Column]`
+  - `#[Column('token_hash', length: 64)]` on `$tokenHash` → `#[Column(length: 64)]`
+  - `#[Column('last_used_at')]` on `$lastUsedAt` → `#[Column]`
+  - `#[Column('expires_at')]` on `$expiresAt` → `#[Column]`
+  - `#[Column('created_at')]` on `$createdAt` → `#[Column]`
+
+- `packages/database/tests/Entity/EntityMetadataFactoryTest.php`:
+  - Line 175: `#[Column(name: 'user_id')]` on `$userId` — this is now redundant. Remove `name:` parameter.
+  - NOTE: The "uses Column attribute name when specified" test reworking is handled in task 001.
+
+- `packages/database/tests/Entity/EntityHydratorTest.php`:
+  - Line 25: `#[Column('email_address')]` on `$email` — NOT redundant (`email` -> `email`, not `email_address`). Keep as-is.
+
+- `packages/database/tests/Repository/RepositoryTest.php`:
+  - Line 32: `#[Column('email_address')]` on `$email` — NOT redundant. Keep as-is.
+
+**Tests to update (check Column attribute `name` via reflection):**
+
+- `packages/admin-auth/tests/Unit/Entity/AdminUserTest.php`:
+  - Line 67: `expect($rememberTokenColumn->name)->toBe('remember_token')` — after removal, `$columnAttr->name` will be `null`. Update test to check `null` or remove assertion (the factory handles the conversion, not the attribute).
+  - Line 72: `expect($isActiveColumn->name)->toBe('is_active')` — same issue. After removal of explicit name from `#[Column(default: '1')]`, `name` is `null`.
+
+- `packages/admin-auth/tests/Unit/Entity/PermissionTest.php`:
+  - Line 117: `expect($columnAttribute->name)->toBe('created_at')` — same issue.
+
+- `packages/admin-auth/tests/Unit/Entity/RoleTest.php`:
+  - Lines 151, 165: `expect($columnAttribute->name)->toBe('created_at')` and `'updated_at'` — same issue.
+
+- `packages/authentication-token/tests/Entity/PersonalAccessTokenTest.php`:
+  - Line 38: `$tokenableTypeColumn->name->toBe('tokenable_type')` — will be `null` after removal. Update to `->toBeNull()` or remove.
+  - Line 45: `$tokenableIdColumn->name->toBe('tokenable_id')` — same issue.
+  - Line 57: `$tokenHashColumn->name->toBe('token_hash')` — same issue.
+  - Line 72: `$lastUsedAtColumn->name->toBe('last_used_at')` — same issue.
+  - Line 79: `$expiresAtColumn->name->toBe('expires_at')` — same issue.
+  - Line 86: `$createdAtColumn->name->toBe('created_at')` — same issue.
+
+- `packages/media/tests/Entity/MediaTest.php`:
+  - Line 45: `$originalFilenameColumn->name->toBe('original_filename')` — will be `null` after removal. Update to `->toBeNull()` or remove.
+  - Line 53: `$mimeTypeColumn->name->toBe('mime_type')` — same issue.
+  - Line 88: `$createdAtColumn->name->toBe('created_at')` — same issue.
+  - Line 96: `$updatedAtColumn->name->toBe('updated_at')` — same issue.
+
+## Requirements (Test Descriptions)
+- [ ] `it removes redundant explicit Column name from admin-auth entities`
+- [ ] `it removes redundant explicit Column name from media entities`
+- [ ] `it removes redundant explicit Column name from webhook entity`
+- [ ] `it removes redundant explicit Column name from authentication-token entity`
+- [ ] `it updates admin-auth entity tests to not assert explicit Column attribute name`
+- [ ] `it updates authentication-token entity test to not assert explicit Column attribute name`
+- [ ] `it updates media entity test to not assert explicit Column attribute name`
+- [ ] `it passes all tests across affected packages after cleanup`
+
+## Acceptance Criteria
+- No redundant explicit `name:` parameters remain in production entity `#[Column]` attributes
+- Admin-auth entity tests updated to reflect that Column attribute `name` is now `null` (auto-derived by factory)
+- Authentication-token entity test updated to reflect that Column attribute `name` is now `null` (auto-derived by factory)
+- Media entity test updated to reflect that Column attribute `name` is now `null` (auto-derived by factory)
+- All tests pass across: database, admin-auth, media, webhook, authentication-token, notification-database packages
+
+## Implementation Notes
+(Left blank - filled in by programmer during implementation)