What problem does it solve? After implementing a package change, it is easy to either under-validate (missing regressions, stale generated artifacts, broken package checks) or over-validate (absorbing unrelated failures and cleanup into the PR). This Skill provides a disciplined final verification gate that proves the intended change works while keeping the PR within its original scope. ## Core Features & Use Cases - Ordered validation sequence: Runs focused evidence first, then neighboring tests, the full test suite, documentation/build checks, package checks, and ecosystem-specific checks such as R CMD check or BiocCheck. - Failure attribution framework: Classifies every failure as introduced, exposed, pre-existing, environmental, or unclear, and fixes only what the intended change owns. - Scope and hygiene audit: Reviews the final diff, generated artifacts, and repository cleanliness so validation never expands the PR's semantic scope. - Use Case: After fixing a bug in an R package, use this Skill to re-run the regression test, run the full suite, rebuild documentation, run R CMD check, attribute any remaining failures, and produce a validation report before requesting review. ## Quick Start Use the validate-a-package-change skill to verify my package change before I request review, starting from the focused regression test and ending with a full validation report.
