Stable diffusion is a generative algorithm for creating images from text prompts. This paper explores prompts with color terms and proposes a process to generate, visualize and assess these results. Automated prompts are used to generate and render a color term, an object and a context. The results are then evaluated using two dashboard views of the underlying images. First is a sampling based on a collection of frequently used color terms. Second is a sampling by object prompts, such as apples and boxes. This paper considers the following questions: how effectively are the colored objects generated? how do the colors generated by stable diffusion compare to human color naming? How might color terms be useful in visualizing properties and features of generative algorithms? The dashboard view of color terms suggests that less frequently used color terms may be generated less consistently. In addition, even the most common color terms can fail to be correctly generated. Likewise, objects with more frequent color associations, such as apples or pumpkins, will result in less accurate color generation.