kangarooking/design-skills
Productivity workflowThis page was translated automatically. Technical details may differ from the original. View the original
Design Skills: Open-source Agent Skills for AI Image Generation and Visual Design | Design Skills: Open-source Agent Skills for AI image creation and visual design.
Overview
Design Skills: Incorporate color schemes, composition, character consistency, and delivery processes into the Agent, making raw image design more organized. It is an open-source Agent Skills collection for AI image generation and visual design. Each skill focuses on specific design tasks, organizing requirements understanding, aesthetic decision-making, prompts, image generation, and result verification into reusable workflows. Subsequent design skills for posters, brand visuals, IP image, and more will continue to be included here. It can be used in agents such as WorkBuddy, Claude Code, Codex, etc., that support loading SKILL.md. Actual image generation requires the current environment to provide corresponding tools; Skill itself does not include model services, interface permissions, or free generation quotas. Currently includes 2 design skills: Halloween poster advanced color scheme (V0.2) and 3D blind box trendy toy IP setting images, both available. The author confirmed the completion of testing on 2026-10-20; Upon inclusion, formatting, file references, and blind box script smoke verification were also completed. This is the "Halloween Poster Advanced Color Technique." It offers four color strategies: accent color, theme color, adjacent color, and orange-blue complementary scheme, and also supports a custom mode prioritizing brand colors. Decide on color schemes, visual styles, and layouts separately, clarify HEX anchor points, area suggestions, information levels, and text contrast, and deliver proposals, prompts, images, or review comments as needed. Suitable for holiday events, beverage products, party invitations, and main visual operations. When accurate Chinese, logos, or event information are needed, base images can be created and then deterministic layout. Color values and scales are design constraints, not guarantees for precise model execution. This is the "3D Blind Box Trendy Toy Concept Skill." Using character reference images as identity anchors, create precise 1:2.5 character design long images, organizing five sections: Hero Banner, Character Profile, Four-View View, Action Banner, and Action Image Collection. Chinese typesetting requires usable Chinese fonts. The script will look for system fonts; if it cannot find a suitable font, it must specify the font file via --font / --font-bold. Generated dimensions are subject to the actual tool support; final proportions must pass pixel checks.
Overview (machine translated)
🎨 Design Skills · Design Skills
Design Skills: Incorporate color schemes, composition, character consistency, and delivery processes into the Agent, making raw image design more organized.
Design Skills is an open-source Agent Skills collection for AI image generation and visual design. Each skill focuses on specific design tasks, organizing requirements understanding, aesthetic decision-making, prompts, image generation, and result checking into reusable workflows. Subsequent design skills such as posters, brand visuals, and IP images will continue to be included here.
Can be used in agents such as WorkBuddy, Claude Code, Codex, etc., that support loading SKILL.md. Actual image generation requires the appropriate tools provided in the current environment; Skill itself does not include model services, interface permissions, or free generation quotas.
Currently includes 2 design skills: Halloween poster advanced color scheme (V0.2) and 3D blind box trendy toy IP setting images, both available. The author confirmed testing completion on 2026-10-20; Upon inclusion, formatting, file references, and blind box script smoke verification were also completed.
📋 Skills
| Name | In a sentence | Status |
|---|---|---|
| 🎃 halloween-poster-design | Choose a color scheme, plan color proportions, composition, and fonts, generate Halloween poster prompts or use available tools to produce images | Available · V0.2 |
| 🧸 blindbox-ip-sheet | Organize characters into five blocks of 3D blind box setting long images, and use scripts to fix action grids, archive text, and single views | Available |
🔎 Skills Description
halloween-poster-design
This is the “Halloween Poster Advanced Color Technique.” It offers four color strategies: accent color, theme color, adjacent color, and orange-blue complementary scheme, and also supports a custom mode prioritizing brand colors. Decide on color schemes, visual styles, and layouts separately, clarify HEX anchor points, area suggestions, information levels, and text contrast, and deliver proposals, prompts, images, or review comments as needed.
Suitable for holiday events, beverage products, party invitations, and main visual operations. When accurate Chinese, logos, or event information are needed, base images can be created and then deterministic layout. Color values and scales are design constraints, not guarantees for precise model execution.
Example request:
blindbox-ip-sheet
This is the “3D Blind Box Trendy Toy Concept Skill.” Using character reference images as identity anchors, create precise 1:2.5 character design long images, organizing five sections: Hero Banner, Character Profile, Four-View View, Action Banner, and Action Image Collection.
Includes white-background cutout, action mesh composition, file text layout, and single-view replacement scripts. Scripts depend on Python 3.10+ and Pillow:
Chinese typesetting requires usable Chinese fonts. The script will look for system fonts; if it cannot find a suitable font, it must specify the font file via --font / --font-bold.
This page was translated automatically. Technical details may differ from the original.
Recommended Tools
Try a different keyword or remove a filter.
Install
npx skillfish add kangarooking/design-skills