design-an-interface

You can explore multiple API designs by spawning 3+ parallel sub-agents to generate radically different interface shapes and compare clear trade-offs for module evaluation.

The best way to compare interface prototypes is by generating multiple design options side-by-side, presenting interface signatures, usage examples, and hidden internal logic to evaluate distinct trade-offs.

You need to generate radically different module boundaries during API design and module experimentation when you must evaluate diverse interface options and contract trade-offs before finalizing an interface.

Yes, this approach works for evaluating API contracts by spawning parallel sub-agents to generate distinct designs, which is useful when deciding on module boundaries and contracts with users.

To enforce design diversity during interface architecture experimentation, coordinate 3+ sub-agents to independently surface diverse shapes, ensuring radically different module designs are generated and compared.

A limitation when generating distinct interface signatures is that design quality relies on sub-agent coordination to enforce diversity, meaning evaluation depends on parallel agents surfacing genuinely distinct trade-offs.