# Project Context: AI Char Bot

## Overview
Python backend project using FastAPI and MongoDB (Motor).
Root: `/Users/xds/develop/py projects/ai-char-bot`

## Architecture
- **API Layer**: `api/endpoints` (FastAPI routers).
- **Service Layer**: `api/service` (Business logic).
- **Data Layer**: `repos` (DAOs and Repositories).
- **Models**: `models` (Domain models) and `api/models` (Request/Response DTOs).
- **Adapters**: `adapters` (External services like S3, Google Gemini).

## Coding Standards & Preferences
- **Type Hinting**: Use `Type | None` instead of `Optional[Type]` (Python 3.10+ style).
- **Async/Await**: Extensive use of `asyncio` and asynchronous DB drivers.
- **Error Handling**:
    - Repositories should return `None` if an entity is not found (e.g., `toggle_like`).
    - Services/Routers handle `HTTPException`.

## Key Features & Implementation Details
- **Generations**:
    - Managed by `GenerationService` and `GenerationRepo`.
    - `toggle_like` returns `bool | None` (True=Liked, False=Unliked, None=Not Found).
    - `get_generations` requires `current_user_id` to correctly calculate `is_liked`.
- **Ideas**:
    - Managed by `IdeaService` and `IdeaRepo`.
    - Can have linked generations.
    - When fetching generations for an idea, ensure `current_user_id` is passed to `GenerationService`.

## Recent Changes
- Refactored `toggle_like` to handle non-existent generations and return `bool | None`.
- Updated `IdeaRouter` to pass `current_user_id` when fetching generations to ensure `is_liked` flag is correct.