cover-image-pipeline

Cover Image Pipeline

What This Skill Does

Generates a complete, coordinated set of image prompts for a cover image concept across multiple formats and crop ratios — so a single visual idea produces consistent, properly composed assets for podcast artwork, YouTube thumbnails, newsletter headers, and social media simultaneously.

When To Use This Skill

• You have a cover image concept and need to generate it in multiple formats (square for podcast, 16:9 for YouTube, vertical for Instagram) without it looking like a crop of the same image
• You are launching a new episode, series, or publication issue and need a full suite of coordinated visual assets from one image session
• You want to ensure that text placement areas, focal points, and compositional choices are optimized for each platform rather than just resized
• You need consistent visual identity across a release campaign that touches multiple platforms simultaneously

What You Need To Provide

Required:

• The core image concept (what the image should show — subject, scene, mood)
• The visual style (or reference to your project's visual identity brief)
• Which formats you need: select from podcast square (1:1), YouTube thumbnail (16:9), Instagram portrait (4:5), Instagram/Facebook landscape (1.91:1), newsletter header (3:1 approximately), or specify custom ratios

Optional:

• Whether the image needs to accommodate text overlay (title, episode number) and where
• The image generation tool being used (Midjourney, Flux, DALL-E) — each has different prompt grammar
• Any existing assets to maintain consistency with (describe or reference your style brief)
• The project's color palette if defined
• Priority format (the one that must be strongest if trade-offs are needed)

How the Assistant Approaches This

1. Analyzes the core image concept and identifies which visual element is the focal point — this determines how each format is composed, not just cropped
2. Designs each format as an intentional composition, not a crop: the 16:9 version may use the wide aspect to show context the square version cannot; the square version may push the focal point to center in a way that would be static in landscape
3. Writes a coordinated prompt set where all versions share a consistent subject, style, and palette anchor, but each prompt is written to optimize for its specific ratio and use context
4. Notes text-safe zones for each format — where title text can be placed without covering the focal point
5. Identifies the generation sequence: which format to generate first (usually the most compositionally complex), and how to use that result's style/seed to anchor the others
6. Closes with a "Next Step" note: confirm the first format generation before moving on to the others, note the seed number after the first successful result, and review the full set at actual display sizes before finalizing