design-systems-slds-validate

Run the SLDS linter with npx @salesforce-ux/slds-linter lint on the component path, then run the analyze-quality.cjs script for supplementary checks. Combine both outputs with a manual review of states and blueprint structure to compute the final grade and recommendation.

The scorecard measures five weighted categories: Linter Compliance at 30%, Theming at 20%, Accessibility at 20%, Code Quality at 15%, and Component Usage at 15%. Each category starts at 100 and loses points per finding, producing an overall A through F grade.

No, this skill only audits and reports violations. Fixing linter violations is handled by the design-systems-slds2-migrate skill, and building new components is handled by design-systems-slds-apply.

Yes, partially. If the linter cannot run, the report notes it as not run, marks Linter Compliance as N/A, and renormalizes the overall score across the remaining four categories. The analysis script itself requires Node.js to execute.

No, the accessibility checks verify attribute presence only, such as labels, alt text, and focus indicators. They do not validate contrast ratios, keyboard flows, or screen reader behavior, so passing does not guarantee WCAG compliance.

The automated grade reflects only linter and script findings. The manual review gate can return Blocking for issues like missing error states or incorrect blueprint structure, which overrides the automated score and marks the component not ready for production.