AgentSkillsCN

illustrator

基于 Google Imagen 4.0 的情境感知型品牌插画系统。该系统能够从当前项目中挖掘品牌元素,自动生成各类视觉资产。支持多种特定宽高比:1:1、9:16、16:9、4:3、3:4。其核心功能仅聚焦于图像创作。

SKILL.md
--- frontmatter
name: illustrator
description: A context-aware brand illustration system using Google Imagen 4.0. It discovers branding from the current project to generate visual assets. Supports specific aspect ratios: 1:1, 9:16, 16:9, 4:3, 3:4. Focuses exclusively on image creation.
license: MIT
metadata:
  author: Francisco Ossian
  version: "2.1"

Illustrator

Brand-consistent image generation via Google Imagen 4.0. It discovers project mood (CSS, configs) to ensure visual harmony.

Usage Triggers

  • Generating images matching project UI/UX.
  • Creating assets for landing pages, blogs, or components.

Workflow

  1. Discovery: Inspect tailwind.config, CSS variables (--primary), and existing assets for "vibe" (glassmorphism, vintage, flat).
  2. Alignment: Clarify Theme (e.g., "Cyberpunk"), Context (Hero, Icon), and Metaphor.
  3. Execution: Run generation script.
    bash
    python .ai/skills/illustrator/scripts/generate.py --prompt "[PROMPT]" --aspect_ratio "[1:1|9:16|16:9|4:3|3:4]" --output "assets/[NAME].jpg"
    
  4. Handover: Use uploader to move assets if needed.

Directory Structure

Constraints

  • Discovery First: Never assume a color palette; always try to discover it.