ray-cover

Generate WeChat and X article covers from a validated visual metaphor brief.

160|18|Updated Jul 13, 2026
One-click install
npx skills add https://github.com/imraywang/rayskills --skill ray-cover-imraywang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ray-cover
Source: https://github.com/imraywang/rayskills/tree/main/skills/ray-cover
Command: npx skills add https://github.com/imraywang/rayskills --skill ray-cover-imraywang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires Pillow, and includes scripts (resource) and references (resource) components.

What problem does it solve? Turning a finished article into platform-ready covers for WeChat, X posts, and X Articles usually means inconsistent visuals, wrong aspect ratios, and broken Chinese typography. This Skill compresses the article's core judgment into one visual metaphor and produces correctly sized covers for each platform. ## Core Features & Use Cases - Metaphor-driven cover generation: Extracts the article's thesis, emotion, and conflict, then selects one approved editorial style (collage, minimal, midcentury, quiet-ink, tool-bridge) instead of generic AI imagery. - Per-platform direct generation: Calls the image model separately for WeChat (2100x900), X (1600x900), and X Article (1600x640) with platform-specific prompts, safe zones, and exact cover text. - Validated briefs and deterministic fallback: Validates every cover brief with brief_check.py, and falls back to compose_cover.py for pixel-exact Chinese typography only when the user explicitly requests it. - Use Case: After finalizing a long-form article, ask for WeChat and X covers; the Skill builds a checked brief, generates each platform's cover independently, verifies thumbnail readability, and saves all assets with a manifest. ## Quick Start Use ray-cover to create WeChat and X covers for my finalized article about creator identity as a platform trust asset.

Frequently Asked Questions about ray-cover

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate WeChat and X covers from one article?▼

Provide the finished article and the Skill extracts its thesis, emotion, and conflict into one visual metaphor, then builds a validated cover brief. It calls the image model separately for each platform with size-specific prompts rather than cropping one image.

What image sizes does it produce for each platform?▼

WeChat covers default to 2100x900 (about 21:9), standard X covers to 1600x900 (16:9), and X Article covers to 1600x640 (5:2). Each platform gets its own generation call with explicit safe zones.

Can it handle Chinese typography on covers?▼

Yes. By default the image model renders Chinese title lines, eyebrow, and subtitle directly from quoted text in the prompt. If the user explicitly requests pixel-exact typography, compose_cover.py overlays text deterministically using CJK fonts.

Does it copy styles like Vox or specific artists?▼

No. Brand or artist names are forbidden in prompts; the brief validator rejects markers like 'vox style'. Instead it uses five transferable editorial style systems describing composition, palette, and material principles.

What happens if the article's core judgment is not ready?▼

The Skill refuses to generate covers and routes back to the writing stage. A stable one-sentence thesis and core conflict are required inputs, so it will not produce generic AI-themed imagery for an undecided article.