Designs and directs the visual assets that carry content — carousels, infographics, quote graphics, diagrams, and social imagery — including the generation prompts where they are AI-produced. Use this to turn a written piece into a visual format, design a carousel or infographic, create social graphics, or fix visuals that are not stopping the scroll.
npx skills add https://github.com/cbrock84/headcount --skill visual-content
once.
Choosing format by what is fashionable rather than by the content's shape is why so many carousels
are a paragraph cut into ten pieces.
The first slide is the entire hook — treat it as a headline, not a title card. Every subsequent
slide must earn the swipe, which means each ends slightly incomplete.
Consistent template across slides: same margins, same type positions, same palette. Variation across
slides reads as error. Number them so the reader knows the commitment.
Last slide lands the idea. A last slide that only says "follow for more" wastes the only slide with
guaranteed attention from everyone who finished.
A diagram earns its place when it shows a relationship prose would take three paragraphs to
establish. If it restates the text, cut it.
Label directly on the element. Legends make the reader hold a mapping in memory while looking
elsewhere. Strip anything not carrying meaning — decorative gradients, 3D effects, and drop shadows
on data.
Almost all of this is seen small, on a phone, in a feed, at speed. Test at that size before
finishing. If the headline is not readable at thumbnail scale, it does not exist.
High contrast between subject and background is what stops the scroll. Subtlety is a desktop luxury.
Write the prompt as a brief: subject, composition, lighting, palette, mood, and negative
constraints. Generate one concept per image. Then check the output actually renders any embedded
text correctly and matches the brand palette — generated imagery drifts, and drifted brand color is
worse than no brand color.
Take cbrock84/visual-content from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.