Skill UI

A comprehensive skill for generating multimodal media content using fal.ai models. It supports text-to-image (Nano Banana), text/image-to-video (Veo, Kling, Seedance), and text/video-to-audio (CSM-1B, ThinkSound). Use this tool for any creative project requiring high-quality generated images, videos, or sound effects.

This skill helps design the complete user journey or 'flow' before building individual screens. It applies established UX principles (like progressive disclosure and hub-and-spoke models) to define entry points, exit paths, and the overall navigation architecture. Use it for complex tasks like onboarding, checkout, or multi-step features to ensure a coherent, logical, and highly intuitive user experience.

This skill acts as an elite art director, generating highly structured, premium, and conversion-aware frontend design references. It is optimized for creating multi-section websites, landing pages, and marketing sites, ensuring consistency and visual appeal across every section. Output is designed to be immediately usable by developers or coding models, adhering to strict composition rules to avoid generic AI aesthetics.

This powerful script uses Nano Banana models to facilitate advanced image creation and manipulation. Users can generate high-quality images from simple text prompts, perform detailed image edits (inpainting/outpainting), or combine multiple images into a single collage. It supports various resolutions and aspect ratios, making it ideal for artists, designers, and anyone requiring advanced visual content generation.

A powerful, provider-agnostic skill dedicated to generating high-quality, single-image assets such as posters, key art, and editorial illustrations. This tool utilizes a structured, multi-step prompt composition workflow (Subject, Lighting, Palette, Camera) to ensure cinematic and artistic consistency, regardless of the underlying AI model (Flux, Imagen, Midjourney). Ideal for designers needing professional-grade visual concepts.

A sophisticated automated loop that uses vision or video models to evaluate subjective aesthetic quality (e.g., timing, color grade, layout). Instead of relying on numeric metrics, the system renders multiple parameterized variants into one artifact, asks the AI to rate them, and suggests optimal values, repeating the process until convergence. Ideal for design iteration where human judgment is the primary success criteria.