diff --git a/data/rag_eval/results/plots/context_precision.png b/data/rag_eval/results/plots/context_precision.png
deleted file mode 100644
index 1d8acc1..0000000
Binary files a/data/rag_eval/results/plots/context_precision.png and /dev/null differ
diff --git a/data/rag_eval/results/plots/context_recall.png b/data/rag_eval/results/plots/context_recall.png
deleted file mode 100644
index d51f262..0000000
Binary files a/data/rag_eval/results/plots/context_recall.png and /dev/null differ
diff --git a/src/agent/prompts.py b/src/agent/prompts.py
index 4e406cf..9b1d57f 100644
--- a/src/agent/prompts.py
+++ b/src/agent/prompts.py
@@ -208,6 +208,10 @@
 4. Some steps of the plan may provide general guidelines without actually providing a command, those steps are excluded from the final plan
 5. **NEVER** provide natural language strings inside "command", textual information such as comments should go in "thought" field"""),
         },
+        'routing': {
+            'system': """""",
+            'user': """"""
+        }
     },
     'mistral': {
         'plan': {
diff --git a/src/main.py b/src/main.py
deleted file mode 100644
index 65557d9..0000000
--- a/src/main.py
+++ /dev/null
@@ -1,65 +0,0 @@
-import requests
-
-from src import upload_knowledge
-from src.agent import Agent
-from src.agent.knowledge import Store
-from src.agent.plan import TaskStatus
-from src.agent.tools import TOOLS
-
-
-def cli_test():
-    """testing Agent"""
-    ollama_model = 'llama3'
-    tools_documentation = '\n'.join([tool.get_documentation() for tool in TOOLS])
-
-    # vector_db = Store()
-    # upload_knowledge('../data/json', vector_db)
-
-    # =================================================================
-    agent = Agent(
-        model=ollama_model,
-        tools_docs=tools_documentation
-    )  # , knowledge_base=vector_db)
-    current_session = 0
-    while True:
-        user_input = input("Enter: ")
-        if user_input == "-1":
-            break
-        elif user_input == "exec":  # execute plan
-            execution = agent.execute_plan(current_session)
-            for output in execution:
-                for i, task_overview in enumerate(output):
-                    print(f'{i+1}. {task_overview}')
-                    if task_overview.status == TaskStatus.DONE:
-                        print(f'Output:\n{task_overview.output}')
-
-        elif user_input.split(" ")[0] == "new":  # create session
-            agent.new_session(int(user_input.split(" ")[1]))
-            current_session = int(user_input.split(" ")[1])
-
-        elif user_input.split(" ")[0] == "save":  # save session
-            agent.save_session(int(user_input.split(" ")[1]))
-
-        elif user_input.split(" ")[0] == "load":  # load session
-            current_session = int(user_input.split(" ")[1])
-            session_history = agent.get_session(current_session)
-            for msg in session_history.messages_to_dict_list():
-                print(f'\n> {msg["role"]}: {msg["content"]}')
-            print('> Plans: ')
-            if session_history.plans is None:
-                print('NaN')
-            else:
-                for plan in session_history.plans:
-                    print(plan)
-
-        elif user_input.split(" ")[0] == "rename":  # rename session
-            agent.rename_session(current_session, user_input.split(" ")[1])
-
-        else:  # query
-            for chunk in agent.query(current_session, user_input, rag=False):
-                print(chunk, end='')
-            print()
-
-
-if __name__ == "__main__":
-    cli_test()