Just an experiment, which turned out decently.
No trigger word, adjust weights past 1.0 to increase the effects.

Simple object prompts like "a fox", "a bear", "a house" will produce isolated drawings, while detailed prompts will produce full images.