Multi-Platform Social Media Design

What It Solves

When generating social media images, models often produce a visually pleasing single frame without solving publishing constraints: Instagram and Reels use different aspect ratios, TikTok has a right-side interaction strip, YouTube thumbnails have a duration badge, and Xiaohongshu favors 3:4 with emoji-heavy headline overlays. If key text lands in overlay zones, in-image copy is mixed with the caption, or carousel pages use inconsistent ratios, the asset needs rework. This skill treats platform specs, safe areas, layout structure, and layered copy as explicit constraints before generation.

How It Works

• Clarify the task: confirm platform, format, engagement goal, audience, and tone before design.
• Plan hook and structure: apply a 3-second first-frame rule, then order carousels as hook, value reinforcement, content blocks, and CTA.
• Validate hard constraints: keep ratios consistent, place subjects in the central safe area, ensure text contrast and shadows, follow the user-specified palette, and avoid rendering buttons, duration badges, resolution labels, or other platform UI.
• Deliver layered copy: keep in-image text short, return the caption separately, and adapt wording for Instagram, TikTok, YouTube, LinkedIn, and Xiaohongshu.
• Iterate and expand: revise color, composition, or style after feedback, or extend a single image into a carousel or multi-platform set.

Boundaries

The skill calls a hosted dLazy API through the dLazy CLI and requires an API key. Local media paths supplied to the workflow are uploaded to dLazy media storage so the model can read them, and result URLs are hosted by dLazy. It is useful for producing publish-ready social visual drafts, but it does not replace brand systems, copyright review, ad strategy, or guarantees about final brand compliance.