Implement Core Agent Logics and LangGraph Workflow

Implement the individual agent logics for your defined roles. For each agent, integrate the OpenAI o3 API for reasoning and decision-making. Develop the LangGraph workflow, including the state definition, node functions for each agent, and conditional edges to manage the flow of information. Crucially, implement the Overseer Agent's logic: it must monitor communication logs, retrieve relevant ethical guidelines from the Haystack RAG system (which queries your Qdrant-based ethical ontology), and identify potential deceptive actions or perspective shifts based on discrepancies between agent statements, actions, and ethical norms. Simulate a 'deceptive' behavior for one agent (e.g., omitting specific information) to test detection.