What problem does it solve? When adding package functionality that connects existing data structures to an established external implementation, developers often over-engineer by reimplementing algorithms, building wrapper frameworks, or hiding preprocessing pipelines. This Skill guides you to build thin domain-specific adapters that keep integration code small while delegating the algorithm to the upstream implementation. ## Core Features & Use Cases - Adapter Design Discipline: Enforces the pattern of selecting data, validating method-specific requirements, aligning samples, delegating to the upstream function, and returning the established result type. - Explicit Contracts: Requires documented alignment, missing-data, argument-forwarding, and result contracts so integration behavior is predictable and testable. - Anti-Pattern Detection: Flags reimplemented algorithms, hidden preprocessing pipelines, unnecessary result wrappers, generic frameworks for single methods, and private upstream calls. - Use Case: You want to run constrained ordination between two assays in a MultiAssayExperiment. The Skill guides you to select the assays, align samples explicitly, transpose at the boundary, call vegan::rda(), and return the standard rda object rather than inventing a new result class. ## Quick Start Ask the AI to design a thin adapter that connects your package's standard container to a specific upstream function, following the prepare-delegate-return pattern.
