A critical feature of situated visualization toolkits is the ability to automatically place visualizations by detecting objects and surfaces in the physical environment. However, the few existing available toolkits lack this feature, which slows down development cycles and makes situated visualizations less practical. In this thesis, I present AVAR-X, a toolkit to create situated visualizations that boosts the agility in development cycles through spatial mapping and object recognition. To this end, I extended AVAR, an existing toolkit, with situating modalities as well as revised the user interface such that development of situated visualizations is user-friendlier. Using AVAR-X, users can situate visualizations by selecting targets in the environment, using two predefined modalities, eliminating the need to do this by hand. To demonstrate the practical application of AVAR-X, I present three usage examples, developed using the toolkit. They illustrate how AVAR-X may be used to quickly situate energy efficiency visualizations in a room, how situated visualizations can help in day-to-day scenarios such as managing the inventory of a supermarket, and how troubleshooting complex computer networks can be made easier. While AVAR-X indeed simplified the development and situating of visualizations, further improvements could be made. Exploring different methods of text entry, or integrating common programming tools, are only a couple of ways of helping the user with agile development of situated visualizations.