fix: align config.yaml structure with Pydantic AppConfig schema

config.yaml nested cheshire_cat and face_detector under the 'services' key,
and llama URLs under 'services.llama'. But AppConfig expects:
- services -> {url, amd_url} (llama endpoints directly)
- cheshire_cat -> top-level key
- face_detector -> top-level key

Because Pydantic silently ignores extra fields, ServicesConfig received
{llama: {...}, cheshire_cat: {...}, face_detector: {...}} and none matched
its 'url'/'amd_url' fields, so ALL service config from YAML was silently
ignored and Pydantic defaults were always used instead.

Flattened services to contain url/amd_url directly, and moved cheshire_cat
and face_detector to top-level keys matching the AppConfig model. Verified
both AppConfig(**yaml_data) and config_manager dot-path traversal work.

bot/config_manager.py

@@ -35,12 +35,15 @@ class ConfigManager:
     def __init__(self, config_path: Optional[str] = None):
         """Initialize configuration manager."""
         self.config_path = Path(config_path) if config_path else Path(__file__).parent.parent / "config.yaml"
-        self.runtime_config_path = Path(__file__).parent.parent / "config_runtime.yaml"
         
         # Memory directory for server configs and state
+        # This directory is volume-mounted in Docker (./bot/memory:/app/memory)
         self.memory_dir = Path(__file__).parent / "memory"
         self.memory_dir.mkdir(exist_ok=True)
         
+        # Runtime config must live inside memory_dir so it persists across container restarts
+        self.runtime_config_path = self.memory_dir / "config_runtime.yaml"
+        
         # Load configurations
         self.static_config: Dict = self._load_static_config()
         self.runtime_config: Dict = self._load_runtime_config()

config.yaml

@@ -6,17 +6,16 @@
 
 # Service Endpoints
 services:
-  llama:
-    url: http://llama-swap:8080
-    amd_url: http://llama-swap-amd:8080
+  url: http://llama-swap:8080
+  amd_url: http://llama-swap-amd:8080
 
-  cheshire_cat:
-    url: http://cheshire-cat:80
-    timeout_seconds: 120
-    enabled: true  # Set to false to disable Cheshire Cat integration
+cheshire_cat:
+  url: http://cheshire-cat:80
+  timeout_seconds: 120
+  enabled: true  # Set to false to disable Cheshire Cat integration
 
-  face_detector:
-    startup_timeout_seconds: 60
+face_detector:
+  startup_timeout_seconds: 60
 
 # AI Models
 models: