Our research explores how humans 'bootstrap' by reusing chunked action sequences to solve complex problems, integrating perspectives from cognitive psychology and problem-solving research. By developing a novel puzzle task, we investigate how participants form and reuse abstract schemas to generalize solutions to new, more complex problems. This work bridges traditionally underrepresented disciplines in cognitive science, such as action sequence learning and transfer learning, while offering a novel interdisciplinary approach to understanding the human mind. Our findings not only demonstrate the human capacity for schema-based problem-solving but also provide new insights into how humans tackle complex tasks with minimal trial-and-error, advancing our understanding of cognitive flexibility and generalization.