Check your web UI against Vercel's official accessibility and design guidelines. Get immediate feedback on semantic HTML, focus states, forms, animations, typography, images, and more.

Many AI-generated interfaces look generic and lack creative character. This skill guides Claude Code to create production-grade web interfaces with exceptional aesthetic quality, distinctive visual identity, and thoughtful design decisions that avoid common AI patterns.

Maintaining consistent brand identity across documents is time consuming and error prone. This skill provides instant access to corporate brand standards including color palettes, typography, formatting rules, and validation tools to ensure all documents meet professional brand guidelines.

Creating Grasshopper definitions manually is time-consuming and repetitive. This skill enables AI assistants to programmatically create components, establish connections, and execute complete parametric modeling workflows through the MCP protocol, turning design specifications into working Grasshopper definitions automatically.

Transform raw user interview data into actionable personas. Create data-backed user profiles with goals, frustrations, behaviors, and design implications in minutes instead of hours.

Creating consistent design tokens manually is time-consuming and error-prone. This tool generates complete design systems including color palettes, typography scales, spacing grids, and component styles from a single brand color input.

Generic AI outputs produce bland, derivative web designs that lack character. This skill guides Claude to commit to bold aesthetic directions and deliver production-grade, memorable frontend code with intentional creative choices.

Build distinctive web interfaces that stand out from generic AI-generated aesthetics. Create memorable, visually striking components with careful attention to typography, motion, and spatial composition.

Video outpainting typically requires complex ComfyUI setups and manual workflow configuration. This skill lets you extend video spatial canvas and change aspect ratios through simple runcomfy CLI commands with AI-powered Wan 2-7 models.

Remove unwanted objects, watermarks, and people from video frames automatically. This skill routes your editing request to the best RunComfy AI model for region-aware video inpainting without requiring manual frame-by-frame work.

Transform existing videos with AI precision. This skill routes your video editing requests to the optimal AI model for restyling, background swaps, motion transfer, or outfit changes while preserving identity.

Creating AI-generated images and videos usually requires navigating complex APIs and writing boilerplate code. The RunComfy CLI skill lets you generate, edit, and transform media by running simple commands from your terminal, with automatic polling and download handling.

Changing image lighting traditionally requires a full reshoot or complex editing software. This skill lets you transform lighting direction, color temperature, and mood using natural language prompts through the RunComfy platform.

Creating lip-synced videos typically requires navigating multiple different AI models with incompatible APIs. This skill picks the right RunComfy model for your input type and executes the correct runcomfy command, supporting portrait-to-avatar, video mouth-swap, and script-to-video workflows.

Kling 3.0 produces multi-shot cinematic video with synchronized audio and consistent characters across scenes. This skill connects Claude Code to six Kling 3.0 endpoints on RunComfy, from affordable Standard 1080p drafts to native 4K masters with audio.

Images are often cropped to the wrong aspect ratio, losing context. This skill extends any image beyond its original boundaries by adding AI-generated content that matches the existing scene, lighting, and style.

Manually editing images to remove unwanted objects or fill missing areas is time-consuming and requires professional tools. This skill lets you perform mask-driven inpainting through the RunComfy CLI, removing watermarks, objects, and blemishes or replacing entire regions with simple text prompts.

Users need to edit images but lack guidance on which AI model to use. This skill routes requests to the optimal model based on intent, from batch processing to multilingual text edits.

Editing images with precise layout control and multilingual text is difficult with generic tools. This skill provides direct access to OpenAI GPT Image 2 edit endpoint with optimized prompting patterns for identity preservation and text editing.

Need precise local edits to images without regenerating the whole scene? Flux Kontext Pro delivers single-reference image editing with high fidelity and strong prompt control, accessible via simple RunComfy CLI commands.