Most of the git repositories I work(ed) on are available here or on my github page.
During the course of my PhD, I was involved in two programming projects related to my research.
Hermipy is a thin Python library that allows automating most of the operations involved in implementing a Hermite spectral method. The library uses the data structures provided by the Numpy library for linear algebra, and it also depends on Sympy for symbolic calculations. Computationally intensive parts of the computations, such as Hermite transforms, are handed over to a C++ compiled component, and Boost is used for efficient exchange of data with Python. For more information, read the documentation.
This is a high-level framework for the numerical solution of the Cahn-Hilliard equation, implementing a number of semi-implicit numerical methods in FreeFem++ and offering users the means to rapidly (i) define the initial and boundary conditions of the problem, (ii) run a particular numerical method with appropriate parameters, and (iii) post-process the solution. Both two and three-dimensional settings are supported, and the geometry of the domain is read in files generated by the Gmsh mesh-generator.
I also wrote utility software that vim or shell users may find useful.