When micro meets macro the adventure starts! We just come back from the wonderful Budapest Summer School. Chris Carroll (lecturer) changed our life (at least twice). The first time was about two years ago when we find out about his super-smart numerical trick- endogenous gridpoints method. It made our code 10 times faster. This time, he shows us  HARK. HARK is a beautiful python library which contains plenty of numerical solution designed for heterogeneous agent problems.  The open-source nature of HARK will make it easier for other researchers to audit and verify new models and methods, and to collaborate on correcting deficiencies when found. As HARK expands to include more canonical models and more tools and utilities, we can all spend less time managing numerical minutiae and more time fretting about identification arguments and data accuracy.