feat!: migrate to OpenAI native tool calls and fix circular deps (#fuck-gemini)

- Fix circular dependencies in agent/tools
- Migrate from custom JSON to OpenAI tool calls format
- Add async streaming (step_stream, complete_stream)
- Simplify prompt system and remove token counting
- Add 5 new API endpoints (/health, /v1/models, /api/memory/*)
- Add 3 new tools (get_torrent_by_index, add_torrent_by_index, set_language)
- Fix all 500 tests and add coverage config (80% threshold)
- Add comprehensive docs (README, pytest guide)

BREAKING: LLM interface changed, memory injection via get_memory()

infrastructure/api/tmdb/dto.py

@@ -1,26 +1,28 @@
 """TMDB Data Transfer Objects."""
+
 from dataclasses import dataclass
-from typing import Optional
 
 
 @dataclass
 class MediaResult:
     """Represents a media search result from TMDB."""
+
     tmdb_id: int
     title: str
     media_type: str  # 'movie' or 'tv'
-    imdb_id: Optional[str] = None
-    overview: Optional[str] = None
-    release_date: Optional[str] = None
-    poster_path: Optional[str] = None
-    vote_average: Optional[float] = None
+    imdb_id: str | None = None
+    overview: str | None = None
+    release_date: str | None = None
+    poster_path: str | None = None
+    vote_average: float | None = None
 
 
 @dataclass
 class ExternalIds:
     """External IDs for a media item."""
-    imdb_id: Optional[str] = None
-    tvdb_id: Optional[int] = None
-    facebook_id: Optional[str] = None
-    instagram_id: Optional[str] = None
-    twitter_id: Optional[str] = None
+
+    imdb_id: str | None = None
+    tvdb_id: int | None = None
+    facebook_id: str | None = None
+    instagram_id: str | None = None
+    twitter_id: str | None = None