designrique/ai-graphic-design-skill
Developer toolsこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
Claude Code skill for creating logos, brand identities, and visual assets using AI — tool selection, prompt engineering, vectorization pipelines, mockups, automation, and IP safety.
概要
A Claude Code skill for creating logos, brand identities, and visual assets using AI — built from real-world production experience, not theory. I'm a . For years I lived in both worlds — crafting brand identities by hand in Illustrator and studying how machines learn to see. When generative AI hit the design industry, I saw the same problem everywhere: designers either rejected AI entirely or surrendered creative control to it, producing what the industry now calls "AI slop" — generic, soulless visuals with no brand coherence. This skill was born from my need to that makes a brand memorable. I needed a system that would let me produce 10x the output while maintaining the mathematical precision of Bezier curves, the intentionality of color theory, and the strategic depth of brand archetypes.
README
以下は原文 README の抜粋です。原文を見る。
AI Graphic Design — Claude Code Skill
A Claude Code skill for creating logos, brand identities, and visual assets using AI — built from real-world production experience, not theory.
Why This Exists
I’m a graphic designer with a Master’s degree in Computer Science. For years I lived in both worlds — crafting brand identities by hand in Illustrator and studying how machines learn to see. When generative AI hit the design industry, I saw the same problem everywhere: designers either rejected AI entirely or surrendered creative control to it, producing what the industry now calls “AI slop” — generic, soulless visuals with no brand coherence.
Neither extreme works.
This skill was born from my need to scale visual identity production without losing the essence and creative control that makes a brand memorable. I needed a system that would let me produce 10x the output while maintaining the mathematical precision of Bezier curves, the intentionality of color theory, and the strategic depth of brand archetypes.
After curating and distilling 46 specialized sources on AI-assisted graphic design — covering tools, prompt engineering, vectorization pipelines, IP law, and Python automation — I compiled everything into a single, actionable skill that any Claude Code user can install and use immediately.
What It Covers
| Section | Content |
|---|---|
| Tool Selection Matrix | 15 AI tools mapped to specific scenarios (Recraft for SVG, Midjourney for ideation, Firefly for IP-safe work, etc.) |
| Briefing Frameworks | RCAO and StoryBrand templates for defining brand identity before any visual generation |
| Prompt Engineering | Structured formulas per tool — Recraft (shape logic + constraints), Midjourney (artistic vocabulary + --sref/–cref/–seed), DALL-E (natural language), Stable Diffusion (LoRAs + ControlNet) |
| End-to-End Workflow | 5 phases: Briefing → Ideation → Vector Creation → Mockups → Delivery |
| Visual Consistency | Style IDs, seed fixing, moodboards, LoRAs, manual curation techniques |
| Vectorization Pipeline | Raster → Smart Upscaling (4 models) → Vectorizer.ai/Chat2SVG → Bezier cleanup (Box Method) |
| Python Automation | py5, vpype, Aspose.SVG, Blender API for batch generation and post-processing |
| Delivery Formats | SVG/EPS/PDF (vector) vs PNG/JPG (raster) with DPI specs for print and digital |
| IP Safety & Ethics | Risk matrix by tool (Firefly vs Midjourney vs Stable Diffusion) + legal rules (BR/US/UK) |
| Professional Mockups | Displacement Maps + Blend Options in Photoshop; AI alternatives (Phygital+, Nano Banana) |
| Anti-patterns | 7 common mistakes designers make with AI and how to avoid them |
| 2025-2026 Competency Roadmap | The skills a designer needs to evolve from executor to “AI Creative Director” |
The Philosophy
“AI tool interfaces become obsolete in months. Fundamentals last decades.”
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add designrique/ai-graphic-design-skill