Implement full comments system: domain entities (Comment, CommentLike),
value objects (CommentContent), use cases (CRUD, like toggle), SQLAlchemy
repository, API v1 endpoints, web UI with comment form and nested replies,
i18n translations (EN/RU/FR/DE), and E2E tests.
Fix nested reply (reply-to-reply) not displaying — the flat reply_comments
dict was only queried for top-level comment IDs, so deeply nested replies
were saved to DB (incrementing comment count) but never rendered. Switch
to a recursive Jinja2 macro that renders any nesting depth.
- PostLike domain entity (post_id, liked_by) with BaseEntity integration
- Post entity: add like_count field (default 0) and to_dict serialization
- PostRepository interface: add get_like, add_like, remove_like methods
- TogglePostLikeUseCase: toggle logic (like → unlike, unlike → like)
- PostResponseDTO/PostResponseSchema: add like_count field
- PostLikeORM model with FK to posts and cascade delete
- SQLAlchemyPostRepository: implement like query/add/remove with ORM mapping
- DI provider registration for TogglePostLikeUseCase
- API endpoint POST /api/v1/posts/{id}/like (auth required)
- Unit tests: PostLike entity, Post.like_count, TogglePostLikeUseCase (7 tests)
- API tests: POST /api/v1/posts/{id}/like (4 tests)
- Test model files: FEATURE_LIKES.md, TEST_MODEL.md updated
- Add AI code generation requirements to AGENTS.md
- Add module-level docstrings to all 46 Python modules
- Add detailed Google-style docstrings to all classes and functions
- Remove all inline comments following self-documenting code principle
- Include Args, Returns, Raises sections in function docstrings
- Add Attributes and Examples sections to class docstrings
