Worked on enhancing the kvcache-ai/sglang repository by improving the native API documentation, specifically for the new v1/score API. Focused on API development and documentation, the work centered around making the scoring and classification workflows more accessible for downstream integrations. Leveraged Python and machine learning concepts to clarify how the API computes token probabilities for given queries and items, supporting easier onboarding for new developers. Emphasized governance-aligned practices to reinforce quality standards and maintainability. No major bugs were reported during this period, allowing for a concentrated effort on documentation-centric improvements that streamline integration and support ongoing API stability.