Developed the Le Chat Conversational Assistant within the asgeirtj/system_prompts_leaks repository, focusing on establishing a policy-driven conversational framework for enhanced user interaction and information retrieval. The work centered on designing explicit language style guide policies and integrating web browsing instructions, providing a structured approach to prompt design and future extensibility. Leveraging skills in Conversational AI, Natural Language Processing, and documentation, the developer delivered the feature using Markdown to ensure clarity and maintainability. This foundation enables safer, more predictable conversational responses and supports scalable governance across related prompts, addressing the need for robust guardrails in conversational assistant development and deployment.