If you're tackling optimization problems with conflicting objectives, this is the go-to Python framework. It's built around evolutionary algorithms like NSGA-II and NSGA-III that find Pareto-optimal trade-offs instead of single answers. The unified minimize() interface works across problem types, from simple single-objective GAs to many-objective problems with 5+ competing goals. Includes benchmark suites (ZDT, DTLZ) for validation and supports constraints, mixed variable types, and parallel evaluation. The ElementwiseProblem base class makes custom problems straightforward to define. Solid for engineering design, resource allocation, or any scenario where you need to explore the solution space rather than pick one metric to optimize.
npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill pymoo