feat: implemented agent mode

agent.py

@@ -6,6 +6,7 @@ from cerebras.cloud.sdk import Cerebras
 
 load_dotenv()
 
+SAMPLE_MODEL = "gpt-oss-120b"
 
 class HanBasicAgent:
     def __init__(self):
@@ -16,7 +17,7 @@ class HanBasicAgent:
         self.tavily = TavilyClient(api_key=self.tavily_key)
         # self.llm_client = OpenAI(base_url="https://router.huggingface.co/v1", api_key=self.hf_token)
         self.llm_client = Cerebras(api_key=os.environ.get("CEREBRAS_API_KEY"))
-        self.model = "gpt-oss-120b"
+        self.model = os.getenv("LLM_MODEL", SAMPLE_MODEL)
 
     def _truncate_query(self, query: str, max_len: int = 390) -> str:
         return query[:max_len] + ("..." if len(query) > max_len else "")