hig-doctor-audit
Audit interface source with HIG Doctor when the user asks to scan, lint, review, or remediate Swift, UIKit, AppKit, React, Vue, Svelte, Angular, Compose, Android XML, Flutter, CSS, or HTML for HIG-ali
SKILL.mdplugin/skills/hig-doctor-audit/SKILL.md
GitHubHIG Doctor audit
Use HIG Doctor findings as review evidence, not certification.
Audit workflow
- Confirm the target repository or directory. Keep the scan within that scope.
- Run
npx -y hig-doctor@2.0.3 <target> --jsonfor machine-readable findings. - Report the detected frameworks, severity counts, rule IDs, locations, and material analysis limits.
- Use
hig_explain_findingor the relevant bundled HIG skill when a finding needs its cited guidance and context. - Before editing, distinguish mechanical fixes from changes that require product, design, accessibility, or platform judgment.
- If the user authorized remediation, preserve justified suppressions and rerun the audit plus the repository's relevant checks.
Gates and baselines
- Use
--fail-on <critical|serious|moderate>only when the requested workflow needs a gate. - Use
--write-baselineonly when the user asks to adopt a baseline. A baseline records accepted existing findings; it does not resolve them. - Do not delete or weaken configuration, inline suppressions, or baselines merely to produce a clean result.
Failure handling
If npx or the package cannot run, state the failed command and the missing
prerequisite. Ask the user to install a supported Node.js runtime or restore
package-registry access. Do not substitute an unversioned audit command.
Output
Lead with the audit result. Separate confirmed findings, judgment-dependent warnings, preserved exceptions, checks run, and missing evidence. Never claim that an automated scan proves HIG compliance or accessibility.