In August 2025, Liuli Hao developed the Le Chat Conversational Assistant for the asgeirtj/system_prompts_leaks repository, focusing on enhancing user interaction and information retrieval through policy-driven prompt design. Leveraging skills in Conversational AI, Natural Language Processing, and documentation, Liuli established a scalable framework that incorporates explicit language style guide policies and web browsing instructions. The assistant was delivered via a commit-based workflow using Markdown, providing a foundation for safer and more predictable conversational responses. This work addressed the need for structured, extensible guardrails in prompt engineering, supporting future governance and extensibility across related conversational AI prompts within the repository.