fix: explicitly disable thinking mode in Qwen InferenceClient calls for both answer and wiki query optimization

agent.py

@@ -71,7 +71,8 @@ class AIBrain:
             completion = self.client.chat.completions.create(
                 model=self.model_name,
                 messages=messages,
-                max_tokens=max_tokens
+                max_tokens=max_tokens,
+                enable_thinking=False
             )
             return completion.choices[0].message.content
         except Exception as e:

utils/wiki_tool.py

@@ -133,7 +133,8 @@ Question: {question}
         completion = ai_client.chat.completions.create(
             model="Qwen/Qwen3-8B",
             messages=[{"role": "user", "content": prompt}],
-            max_tokens=32
+            max_tokens=32,
+            enable_thinking=False
         )
         query = completion.choices[0].message.content.strip()
         # Nếu AI trả về rỗng, fallback