feat: add reasoning_effort config to enable LLM thinking mode

nanobot/providers/base.py

@@ -88,6 +88,7 @@ class LLMProvider(ABC):
         model: str | None = None,
         max_tokens: int = 4096,
         temperature: float = 0.7,
+        reasoning_effort: str | None = None,
     ) -> LLMResponse:
         """
         Send a chat completion request.