Claude Skills by ADWilkinson

Opinionated constraints for building better interfaces with agents.

Use when user requests implementation work (implement, add, create, build, refactor, fix) AND the request lacks clear acceptance criteria, scope, or constraints. Do NOT use during exploration, explanation, or continuation of ongoing work.

Audit UI code for accessibility violations and visual consistency issues with scoring

Remove AI-generated slop from diffs. Includes confidence scoring and false positive lists.

Detect project type and set up appropriate pre-commit hooks

Run Lighthouse audits and iteratively fix issues until target scores are met. Optimizes performance, accessibility, best practices, and SEO.

Manage Linear tasks and issues. Use when the user mentions Linear, tasks, issues, tickets, backlog, sprints, or wants to create/view/update work items.

Ruthlessly simplify interfaces through systematic reduction. Use when UI feels cluttered, has competing elements, or needs production-ready polish. Removes before adding. Less is more.

Autonomous repository improvement. Fire-and-forget command that finds and fixes issues, commits changes to a branch, and raises a PR. No approval needed - the PR is the review gate.

Creates or updates CLAUDE.md files throughout the project to provide comprehensive AI development context. Use this to systematically document your codebase for AI assistants.

Verifies that changes work correctly by running appropriate tests, builds, and checks. Use after implementing features or fixing bugs to ensure everything works. Provides the verification feedback loop that dramatically improves code quality.

Convert prompts to XML format for structured Claude interactions