Skill UI

Skill for building powerful frontend claude.ai artifacts using React 18, TypeScript, Vite, Parcel, Tailwind CSS and shadcn/ui. Provides scripts to initialize, develop, bundle and test artifacts, with guidance to avoid common AI slop design patterns.

Build a cloud, SLO, and incident-readiness register after intake. Use when an SME needs monitoring scope, alert ownership, cost limits, or service planning.

Guides investigating common or slow process variants, cases with repeated work, and exact-case queries in Process Mining. Provides step-by-step tool calls for variant metrics, rework filtering, and attribute-based case selection, adhering to shared contracts and investigation methods.

Turns AI agents into email marketing operators: build automation flows and segments, de-slop copy, direct email design, triage deliverability, and drive an ESP via MCP behind hard pre-send gates and compliance checks.

Build or fix frontends to look designed, not AI-generated. Includes taste direction, craft floor checklists, and browser-based quality gate that renders pages at phone and desktop widths, catching overlap, low contrast, and AI-slop defaults. Proof before calling work done.

A step-by-step workflow for UI work: understand the project, choose the right plan (new screen, dashboard, polish, simplify, review), then render, fix, and re-check until the result looks right. It helps agents build real, polished interfaces and avoid AI slop.

This skill enhances frontend design quality by avoiding generic AI defaults. It guides the model to infer user intent, audience, and vibe before coding. Users configure three dials (Variance, Motion, Density) to control aesthetics. It maps briefs to official design systems like Fluent UI or Material. Ideal for landing pages, portfolios, and redesigns requiring specific visual identity.

Audit a pytest suite for surviving mutants, untested code, flaky and order-dependent tests, slow tests, and removal candidates, with report-only findings.

Reviews existing test suites for structural problems like brittleness, mock abuse, coverage illusions, slow runs, and poor readability, drawing on twelve classic engineering books such as xUnit Test Patterns and Working Effectively with Legacy Code.