skills.rest
itwanger
itwangerCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

video-cover-image

Generate matched 3:4, 16:9, and 4:3 video cover images from technical scripts.

17.5k|2.5k|Updated Nov 8, 2018
Search Tags:#image generation#pillow#thumbnail generation#short video#video cover#aspect ratio#chinese typography
One-click install
npx skills add https://github.com/itwanger/toBeBetterJavaer --skill video-cover-image
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: video-cover-image
Source: https://github.com/itwanger/toBeBetterJavaer/tree/main/.agents/skills/video-cover-image
Command: npx skills add https://github.com/itwanger/toBeBetterJavaer --skill video-cover-image

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Creating consistent, platform-ready cover images for short technical videos across multiple aspect ratios is repetitive and error-prone, especially when Chinese text must render exactly and a fixed presenter character must stay visually identical across every cover.

Core Features & Use Cases

  • Matched Multi-Ratio Cover Sets: Produces a coordinated 3:4 vertical cover (小红书/抖音/快手), a 16:9 widescreen cover (B站), and a 4:3 horizontal cover from a single video script.
  • Two Cover Modes: A default illustrated mode with a fixed 二哥 cartoon presenter and blue-sky knowledge-cover style, plus a pure-text mode with black/gray background, white title, and yellow keyword lines rendered deterministically via a Python script.
  • Script-Driven Extraction: Reads Markdown scripts under docs/src/ai/video/ to extract the topic, a question-style main title, and a visual metaphor for the cover.
  • Use Case: Given a script about Claude Code's short-term memory, generate three covers titled "Claude Code 的短期记忆是如何实现的？" with the same presenter, color system, and framing across all ratios.

Quick Start

Use the video-cover-image skill to generate 3:4, 16:9, and 4:3 covers from the script docs/src/ai/video/claude-code-short-term-memory.md.

Frequently Asked Questions about video-cover-image

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

FAQPage Schema
How do I generate video cover images in multiple aspect ratios?▼

Provide one video script or topic and the skill generates a matched set of 3:4 vertical, 16:9 widescreen, and 4:3 horizontal covers. All three share the same title, visual cue, character, and color system, with each canvas recomposed rather than mechanically cropped.

How do I create a pure-text cover with white title and yellow keywords?▼

Request a pure-text or 纯文本 cover, or provide a black/gray reference card. The skill runs scripts/render_text_cover.py with a title and one or two keyword lines, deterministically rendering all three ratios with a charcoal textured background using Pillow.

Can I generate covers for B站, 抖音, and 小红书 at the same time?▼

Yes, that is the default behavior. The 3:4 cover targets 小红书, 抖音, and 快手 mobile feeds, the 16:9 cover targets B站 widescreen thumbnails, and the 4:3 cover serves platforms preferring a less-wide thumbnail.

Does the skill keep the same presenter character across covers?▼

Yes, every illustrated cover uses references/itwanger-cover-character.png as the canonical presenter reference, preserving the black spiky hair, large black glasses, brown eyes, yellow shirt, and black tie. Only expression, gaze, and pose adapt to the topic.

Why does Chinese text sometimes garble in generated covers?▼

Image generation models can distort Chinese glyphs in dense or long text. The skill mitigates this by keeping one short question-style title, and for exact wording it uses the deterministic Pillow rendering script instead of generative text.

When should I use the pure-text mode instead of the illustrated mode?▼

Use pure-text mode when the user asks for a text-only cover, provides a black/gray reference card, or wants the 白色大字+黄色小字 layout. Use illustrated mode for normal knowledge-zone covers needing the presenter, icon, and blue-sky style.

Related Skills

Explore More Content & Communication Skills→
ormastesormastes

browser_research

Collect browser-based domain research into markdown with screenshots and extracted data.

Community
Intermediate
momozi1996momozi1996
519

qiuzhi2046-skill

Convert AI user pain points into step-by-step tutorials and One-Click workflows.

Community
Intermediate
JlsilesJlsiles

hook-writing

Identify a video's compelling moment and write a 30-second hook across narration, visuals, and text overlays.

Community
Advanced
BrandondaymdrBrandondaymdr
1

photoshop-thumbnails

Create YouTube thumbnails in Adobe Photoshop with brand-aligned templates.

Community
Intermediate