How to have conditional loops in multiagent framework?

[asokans11]

I am trying to devise a workflow where one of the agent node's job is to verify the user. In the examples that langchain provided the node returned directly to the supervisor but what i am trying to do is to return to the supervisor only if verified. My current code has unnecessary self loops and a deterministic edge to the verification node. It so happens because of how I defined my conditional edge. Ill attach my Supervisor and Graph builder code below and the architecture I am trying to achieve with this. Thank you
Code:
Supervisor.py
def supervisor_node(state: OverallState) -> Command[Literal["qa", "verification", "end"]]:
messages = [
{"role": "system", "content": system_prompt},
] + state["messages"]
response = llm.with_structured_output(Router).invoke(messages)
print(response)
goto = response["next"]
if goto == "FINISH":
goto = END

return Command(goto=goto, update={"next": goto})
graph.py:
def verified(state: OverallState)-> Literal["supervisor", "verification"]:
verified_successfully = state["verified"]

if verified_successfully:
    return "supervisor"
else:
    return "verification"

chatbot = StateGraph(OverallState)
#Adding Nodes
chatbot.add_edge(START, "supervisor")
chatbot.add_node("supervisor", supervisor_node)
chatbot.add_node("qa",qa_node)
chatbot.add_node("verification", verification_assistant)
chatbot.add_node('collect_info', collect_info)
chatbot.add_node("verify_information",verify_information)

#Adding Edges
chatbot.add_edge("verification", "collect_info")
chatbot.add_edge("collect_info", "verify_information")
chatbot.add_conditional_edges("verify_information", verified, "supervisor" , "verification") # problem is with this line

graph = chatbot.compile()
graph.get_graph().draw_mermaid_png(output_file_path="graph.png")
Current Graph:

Graph I am trying to achieve: