Worked on the deepmodeling/abacus-develop repository to deliver a Makov-Payne correction for electrostatic energy calculations in cubic lattices, targeting improved accuracy for isolated systems. The implementation involved integrating the correction directly into the core energy calculation path, allowing for user-defined parameterization to support per-system tuning. C++ was used extensively, leveraging numerical methods and scientific computing techniques to ensure robust and configurable results. The work also included updating the build configuration and unit tests, as well as refining the test suite dependencies and pull request workflow to stabilize energy unit tests and streamline the development process for future contributions.