Brand Style Social Card Designer

Problem

When copy, short-video scripts, or report highlights are finalized but still need to become a mobile-readable social media card set, the hard part is not adding captions. It is deciding page structure, visual style, typography, and rendering constraints. Naive AI card generation often crams long text into one image, swaps components page by page, drifts away from a unified brand system, or drops uncut AI-generated figures directly into layout.

How It Works

The skill turns the job into a controlled pipeline. It first detects image-generation capability and chooses a pure HTML mode or a hybrid mode, then compresses source material into core claim, reader promise, single-point pages, evidence, and action. It plans 4-8 cards with clear responsibilities, recommends or accepts one of 77 built-in brand styles, and locks canvas, color, typography, radius, and component rules before generating pages. In hybrid mode, it generates transparent or chroma-key assets, removes the background, and integrates the result as a layout element rather than a rectangular photo insert. The final output is rendered to high-resolution PNG, then checked for line breaks, density bands, bottom whitespace, overflow, and style consistency, with HTML, PNG, and a contact sheet delivered.

Boundaries

It is useful for carousels, knowledge cards, report summaries, method frameworks, and action steps. It is not intended for copy-only work, single illustration posters, or exact complex charts. Users should provide clear copy or material that can be split into pages, and image sources should be supplied, searched from free libraries, or generated with care because background cleanup and visual relevance affect publication quality.