During February 2025, Cacate Chen extended the differentiation capabilities of the chalk-lab/Mooncake.jl repository by implementing a backward rule for the BLAS.nrm2 operation. This addition enabled robust reverse-mode automatic differentiation for nrm2, improving the accuracy and flexibility of gradient-based workflows within the library. The work involved updating BLAS rule definitions, enhancing associated tests, and preparing the codebase for a version increment to support future releases. Utilizing Julia programming, linear algebra, and numerical differentiation expertise, Cacate focused on feature delivery and release readiness, demonstrating depth in both technical implementation and alignment of documentation with the evolving codebase.