feat(bookslm): v2.2.0 - BooksLM chat AI contextuel par répertoire + 4 providers AI

BooksLM (#76):
- Panneau chat slide-in 450px, clic-droit répertoire → 🧠 BooksLM
- Collecte récursive .md avec limites (200 fichiers, 200K chars)
- Cache SHA-256, redaction secrets, priorité README/index
- SSE streaming, badges sources cliquables, historique localStorage
- Commandes palette: BooksLM ouvrir/nouvelle conversation

Providers AI (4 nouveaux):
- NVIDIA (integrate.api.nvidia.com)
- QwenCloud (dashscope.aliyuncs.com)
- Xiaomi (api.xiaomi.com)
- Mistral (api.mistral.ai)
- Tous OpenAI-compatible, auto-listing modèles

Fix: dropdown config-select suit maintenant le thème (option bg/color)
27 tests BooksLM + 466 tests au total

backend/ai.py

@@ -1,6 +1,6 @@
 """ObsiGate AI — Multi-provider AI service for editor enhancement.
 
-Supports: DeepSeek, OpenRouter, Google Gemini.
+Supports: DeepSeek, OpenRouter, Google Gemini, Ollama, NVIDIA, QwenCloud, Xiaomi, Mistral.
 Configured via environment variables.
 """
 
@@ -14,7 +14,7 @@ import httpx
 
 logger = logging.getLogger("obsigate.ai")
 
-ProviderName = Literal["deepseek", "openrouter", "gemini", "ollama"]
+ProviderName = Literal["deepseek", "openrouter", "gemini", "ollama", "nvidia", "qwencloud", "xiaomi", "mistral"]
 
 # Provider configurations — keys loaded from file or .env
 AI_KEYS_FILE = Path("data/api_keys.json")
@@ -61,6 +61,30 @@ def _load_provider_keys():
             "model": os.getenv("OLLAMA_MODEL", "qwen2.5-coder:1.5b"),
             "auth_header": "Bearer {api_key}",
         },
+        "nvidia": {
+            "api_key": get_ai_key("NVIDIA_API_KEY"),
+            "base_url": "https://integrate.api.nvidia.com/v1",
+            "model": os.getenv("NVIDIA_MODEL", "meta/llama-3.1-405b-instruct"),
+            "auth_header": "Bearer {api_key}",
+        },
+        "qwencloud": {
+            "api_key": get_ai_key("QWENCLOUD_API_KEY"),
+            "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
+            "model": os.getenv("QWENCLOUD_MODEL", "qwen-max"),
+            "auth_header": "Bearer {api_key}",
+        },
+        "xiaomi": {
+            "api_key": get_ai_key("XIAOMI_API_KEY"),
+            "base_url": "https://api.xiaomi.com/v1",
+            "model": os.getenv("XIAOMI_MODEL", "mimo-v2-pro"),
+            "auth_header": "Bearer {api_key}",
+        },
+        "mistral": {
+            "api_key": get_ai_key("MISTRAL_API_KEY"),
+            "base_url": "https://api.mistral.ai/v1",
+            "model": os.getenv("MISTRAL_MODEL", "mistral-large-latest"),
+            "auth_header": "Bearer {api_key}",
+        },
     }
 
 PROVIDERS = _load_provider_keys()
@@ -139,6 +163,7 @@ async def ai_complete(prompt: str, provider: ProviderName | None = None) -> str:
     cfg = _get_provider_config(provider)
     if cfg["name"] == "gemini":
         return await _call_gemini(prompt, "You are a helpful assistant.")
+    # All other providers use OpenAI-compatible format
     return await _call_deepseek_openrouter(prompt, "You are a helpful assistant.", provider)