Motivated by the ever-increasing installation of photovoltaics in homes and small businesses, we consider a general small-scale market for agent-to-agent (energy) resource sharing, in which each agent could either be a producer (server) or a consumer (client) in each time period. We use a simple mechanism under which servers and clients are matched at random, and may each choose an offered price. If the client offers more than the server proposes, the transaction is successful. We model this system in the form of a mean field game, and prove the existence of a mean field equilibrium in which the server's bid is a fixed price that the client either accepts or rejects based on its current budgetary surplus. We also show in simulations that the system can attain near 100% trade ratios. Finally, using a case study over a synthetic electric grid, we show that the system can benefit both individuals and the system as a whole.