Find the capability. Inspect the requirements.

Design process + taste for one-off HTML artifacts. Scopes briefs, produces variants, verifies local HTML artifacts, avoids AI-design slop. Use when the user wants a from-scratch designed artifact with no specific brand dictated.

54 ready-to-paste design systems capturing exact color palettes, typography, component styles, spacing, shadows, and responsive behavior. Each template includes implementation notes, font substitutions, and CSS values. Use when the user wants to build a page styled after a known brand.

Anti-slop frontend skill for landing pages, portfolios, and redesigns. Infers design direction from briefs, applies calibrated dials (variance, motion, density), and ships interfaces that feel designed, not templated. Includes premium-consumer palette rotation, real design system mapping, strict anti-patterns checklist.

Elite frontend image-direction skill for generating premium, conversion-aware website design reference images. One separate horizontal image per section. Enforces composition variety, background-image freedom, and narrative concept spine. Optimized for landing pages and marketing sites developers can accurately recreate.

Frontend design patterns combining LobeHub's premium aesthetic (glass morphism, soft surfaces, generous whitespace) with shadcn/ui's utility-first approach. Covers landing pages, SaaS UIs, dashboards, and product interfaces. Includes responsive breakpoints, icon guidelines, and React/Next.js implementation rules.

Semantic Design System skill for Google Stitch. Generates agent-friendly DESIGN.md files with strict typography, calibrated color, asymmetric layouts, perpetual micro-motion, and hardware-accelerated performance. Includes comprehensive anti-patterns list. Use when generating screens via Google Stitch or Stitch MCP.

Elite mobile app image-direction skill for premium, app-native screen concepts. Prioritizes clean hierarchy, multi-screen consistency, controlled color palettes, and creative art direction. Defaults to showing screens inside a premium iPhone mockup. Use when designing onboarding flows, dashboards, or app concepts.

Edit PDFs using natural-language instructions via the nano-pdf CLI. Point it at a page number and describe what to change: fixing typos, updating dates, changing titles. Supports batched edits. Requires a platform API key for image editing under the hood.

A research harness that breaks a question into sub-queries, runs parallel web searches, fetches and reads sources, adversarially verifies claims, and synthesizes a cited report. Use when you need a multi-source, fact-checked answer rather than a single quick search.

Checks meta tags, heading hierarchy, JSON-LD structured data, internal links, image alt text, and llms.txt coverage, then returns a prioritized fix list. Optimizes for both classic search and AI answer engines. Use before publishing a page or when traffic drops.

Turns a one-line spec into a production-ready React component: typed props, sensible variants, keyboard and screen-reader support, and Tailwind styling that matches your tokens. Ships with usage examples and avoids one-off magic values. Use when building out a component library.

Takes a short brand brief and produces a usable identity starter: color palette with accessible contrast pairs, a type scale, spacing and radius tokens, and a logo direction. Exports tokens as CSS variables or a Tailwind config. Use when starting a new brand or refreshing an old one.

Drafts multi-step email sequences — welcome, onboarding, re-engagement, sales nurture — with subject-line variants and clear calls to action. Matches your tone, avoids spam triggers and corporate filler, and structures each step around one outcome. Use when launching a campaign or automation.

Generates readable charts — line, bar, area, scatter — with sensible axes, labels, tooltips, and color that survives colorblind checks. Picks the right chart for the data and ships responsive, accessible code. Use when turning a dataset into a dashboard or report visual.

Reviews a diff or file for correctness bugs, security issues, performance traps, and reuse opportunities, then returns severity-tagged findings with suggested fixes. Verifies each finding before reporting it to cut false positives. Use before merging to main.

Rewrites copy to be clearer and more human: cuts filler, replaces jargon with plain language, varies sentence structure, and keeps the original intent. Flags overused buzzwords and superlatives. Use on landing pages, product copy, or anything that reads like it was written by committee.