CubeSats operating in a swarm are characterized by a mix of scheduled intermittent connectivity, high delays, and high failure rates. Each CubeSat is limited in size, usually has low data rates and has a low mass. Consequently, they have limited space for solar panels, and thus limit their available energy. Profitably, CubeSats can function in swarms using inter-CubeSat links as well as ground links. Accordingly, any routing protocols developed for CubeSats must be energy aware. We propose two novel Shortest and Energy Reliable Path (SERP) routing protocols; namely, SERP- Breadth-First Search (SERP-BFS) and SERP-Dijkstra. Both algorithms aim to minimize the overall energy cost and maintain connectivity over time. Both choose shortest paths that have CubeSats energy levels higher than or equal to an energy reliability threshold. We have compared our SERP algorithms with Epidemic algorithm. The results show the outperformance of our proposed algorithms in terms of saving the overall energy cost.