fix: use max_completion_tokens for gpt-5.4-nano

gpt-5.4-nano uses max_completion_tokens instead of max_tokens

backend/main.py

@@ -88,7 +88,7 @@ async def run_conversation(text: str, user_id: str) -> str:
             {"role": "system", "content": system_prompt},
             {"role": "user", "content": text},
         ],
-        max_tokens=300,
+        max_completion_tokens=300,
         temperature=0.7,
     )
     kira_text = resp.choices[0].message.content or "Mhm, I'm here!"