Worked on the Embodied-AI-Guide repository to enhance documentation for the Dual-System Hierarchical VLA paradigm, focusing on clear articulation of System 1 fast reaction and System 2 deep thinking components. The approach involved detailed technical writing in Markdown, outlining variations in model structure, communication protocols, and training data, while integrating curated references from both industry and academia. Addressed documentation reliability by fixing a broken external link, ensuring consistent access to supporting materials. The improvements streamlined onboarding and facilitated evidence-based design decisions, supporting faster cross-team collaboration. Emphasized research and documentation skills to deliver comprehensive, accessible guides for the embodied AI community.