The stellar halos of galaxies provide records of their past, including the early epochs of star formation and the accretion of satellites. The clearest record is written in the space of chemodynamics (stellar orbits and abundance patterns), which is very challenging to observe beyond the Milky Way. Instead, globular clusters (GCs) can be used as bright proxies for stellar halos, if their origins and connections with the field stars can be understood. Various fundamental questions about GCs remain unresolved, such as the sites and modes for their formation, and the explanation(s) for their internal variations in stellar populations. In addition, a few peculiar, massive Local Group star clusters (e.g., omega Cen and NGC 2419) have turned out to be representatives of a far-flung family of ultracompact dwarfs (UCDs) whose nature and origins are also unclear. Intriguingly, a new type of stellar system at the other extreme of the density scale has now been recognized: the ultra-diffuse galaxes (UDGs), which show unexpected connections with GCs and with UCDs. I will present new results on masses, orbits, and stellar populations for GCs, UCDs, and UDGs -- based on Keck, Subaru, HST, and Spitzer observations -- and discuss the implications for their origins and for the build-up of stellar halos.