跳转到内容

Brand Guidelines:品牌视觉规范应用

brand-guidelines 是品牌样式的”纯指令型”Skill——通过约 73 行的 SKILL.md 将 Anthropic 官方品牌视觉规范编码为可复用的指令,确保所有 AI 输出严格遵循品牌颜色、字体和视觉风格。

  • 🎨 品牌色彩系统:主色(深色 #141413、浅色 #faf9f5)和强调色(橙色、蓝色、绿色)的精确定义与应用
  • 🔤 字体规范:标题使用 Poppins(回退 Arial)、正文使用 Lora(回退 Georgia)的层级字体策略
  • 🧠 智能字体应用:根据文本层级自动应用不同字体——24pt+ 标题用 Poppins,正文用 Lora
  • 🎯 形状强调色:非文本形状循环使用橙、蓝、绿三色强调,保持品牌一致性的同时增加视觉层次
  • 🔄 自动回退机制:当自定义字体不可用时自动回退到 Arial/Georgia,确保跨系统可读性

当用户涉及品牌颜色、样式指南、视觉格式、企业设计标准、Anthropic 品牌的视觉输出时触发。

brand-guidelines is a “pure-instruction” Skill for brand styling — encoding Anthropic’s official brand visual identity into approximately 73 lines of SKILL.md as reusable instructions, ensuring all AI outputs strictly follow brand colors, typography, and visual style.

  • 🎨 Brand Color System: Precise definition and application of primary colors (Dark #141413, Light #faf9f5) and accent colors (orange, blue, green)
  • 🔤 Typography Specification: Hierarchical font strategy — Poppins for headings (fallback Arial), Lora for body text (fallback Georgia)
  • 🧠 Smart Font Application: Automatically applies fonts based on text hierarchy — Poppins for 24pt+ headings, Lora for body text
  • 🎯 Shape Accent Colors: Non-text shapes cycle through orange, blue, and green accents for brand consistency with visual depth
  • 🔄 Automatic Fallback: Gracefully falls back to Arial/Georgia when custom fonts are unavailable, ensuring cross-system readability

Triggers when users mention brand colors, style guidelines, visual formatting, corporate design standards, or Anthropic brand visual output.

brand-guidelines 是典型的**“纯指令型”** Skill——只有 2 个文件:SKILL.md + LICENSE.txt。全部品牌逻辑都编码在约 73 行的 Markdown 中。

约 73 行的 SKILL.md 分为以下层次:

  • 主色系统:深色/浅色/中灰/浅灰四种主色调的 Hex 值精确定义
  • 强调色板:橙色 #d97757、蓝色 #6a9bcc、绿色 #788c5d——应用于非文本形状
  • 字体系列:Poppins(标题)+ Lora(正文)的经典搭配

SKILL.md 不只是列出字体,而是定义了”如何应用”的规则:

  • 标题(24pt+)→ Poppins
  • 正文 → Lora
  • 智能背景判断 → 自动选择深/浅色
  • 跨平台回退 → Arial / Georgia

定义了强调色的使用策略——循环而非随机选择,确保视觉一致性。

涵盖字体管理、RGB 颜色精确匹配等实现层面的指导。

这个 skill 的核心洞察是:品牌的视觉规范本质上是可编码的规则集合。颜色 Hex 值、字体规则、回退策略——所有这些都可以精确地写入 SKILL.md,让 Claude 在生成任何视觉输出时自动遵循。

与传统的”品牌手册 PDF”不同,将品牌规则编码为 Skill 意味着:

  • Claude 在每次输出时自动应用,无需人工参考手册
  • 品牌一致性从”人工检查”变为”默认行为”
  • 更新品牌规范只需编辑 SKILL.md,无需修改下游流程

brand-guidelines is a typical “pure-instruction” Skill — only 2 files: SKILL.md + LICENSE.txt. All brand logic is encoded in approximately 73 lines of Markdown.

The ~73 line SKILL.md is organized into the following layers:

  • Primary Color System: Precise Hex values for dark/light/mid-gray/light-gray
  • Accent Palette: Orange #d97757, Blue #6a9bcc, Green #788c5d — applied to non-text shapes
  • Font Families: Classic pairing of Poppins (headings) + Lora (body text)

2. Smart Font Application Strategy (~15 lines)

Section titled “2. Smart Font Application Strategy (~15 lines)”

The SKILL.md not only lists fonts but defines “how to apply” rules:

  • Headings (24pt+) → Poppins
  • Body text → Lora
  • Smart background detection → auto selects dark/light color
  • Cross-platform fallback → Arial / Georgia

3. Shape and Accent Color Rules (~10 lines)

Section titled “3. Shape and Accent Color Rules (~10 lines)”

Defines accent color usage strategy — cycling rather than random selection, ensuring visual consistency.

Covers implementation-level guidance on font management, precise RGB color matching, etc.

The core insight of this skill is: brand visual guidelines are essentially a codifiable set of rules. Color hex values, font rules, fallback strategies — all of these can be precisely written into SKILL.md, allowing Claude to automatically follow them in any visual output.

Unlike a traditional “brand manual PDF”, encoding brand rules as a Skill means:

  • Claude automatically applies them on every output, no manual reference needed
  • Brand consistency shifts from “human inspection” to “default behavior”
  • Updating brand guidelines requires only editing SKILL.md, no downstream process changes

brand-guidelines 不包含任何脚本、模板或资产文件

“纯指令”模式在品牌场景中的深度解析

Section titled ““纯指令”模式在品牌场景中的深度解析”

这个 skill 是”纯指令”模式在品牌场景中的典型应用。与需要 Python 脚本处理数据的复杂 skill 不同,品牌引导完全通过指令表达:

为什么品牌规则适合纯指令模式?

Section titled “为什么品牌规则适合纯指令模式?”

品牌规范本质上是一组声明式规则(“标题用这个字体”、“主色是这个”),而非过程式算法(“先解压文件,然后解析 XML”)。Claude 的理解能力足以直接解释和执行这些规则。

品牌 guidlines 转为 Skill 的核心价值是可执行性——PDF 格式的品牌手册需要人工阅读、理解和遵守,而 Skill 格式的品牌指南是 AI 原生可执行的。每次 Claude 输出时,品牌规则自动生效。

这种”品牌规则即指令”的模式很容易推广到任何组织:

  1. 提取品牌核心元素(颜色、字体、Logo 使用规则)
  2. 用精确的 Hex/L 值编码颜色
  3. 定义层级化的字体应用规则
  4. 指定回退策略
  5. 加入格式规则(间距、对齐、留白)

brand-guidelines contains no scripts, templates, or asset files.

Deep Analysis of the “Pure Instruction” Pattern in Brand Context

Section titled “Deep Analysis of the “Pure Instruction” Pattern in Brand Context”

This skill represents the “pure instruction” pattern applied to brand scenarios. Unlike complex skills requiring Python scripts for data processing, brand guidance is expressed entirely through instructions:

Why Are Brand Rules a Good Fit for Pure Instruction?

Section titled “Why Are Brand Rules a Good Fit for Pure Instruction?”

Brand guidelines are essentially a set of declarative rules (“headings use this font”, “primary color is this”), not procedural algorithms (“first unzip the file, then parse XML”). Claude’s comprehension ability is sufficient to directly interpret and execute these rules.

The core value of converting brand guidelines into a Skill is executability — a PDF brand manual requires human reading, understanding, and compliance, while a Skill-format brand guide is AI-natively executable. Brand rules automatically take effect with every Claude output.

This “brand rules as instructions” pattern is easily generalizable to any organization:

  1. Extract core brand elements (colors, fonts, logo usage rules)
  2. Encode colors with precise Hex/RGB values
  3. Define hierarchical font application rules
  4. Specify fallback strategies
  5. Add formatting rules (spacing, alignment, whitespace)
为什么不需要脚本说明
规则是声明式的品牌规则是"如果 X 则用 Y",不需要计算逻辑
输出是直接的视觉元素Claude 直接在生成代码/文档时应用样式规则
回退策略内置字体不可用时的回退已在 SKILL.md 中定义
零依赖不需要 Python、Node.js 或任何外部工具
Why No Scripts NeededExplanation
Rules are declarativeBrand rules are "if X then use Y", no computation needed
Output is direct visual elementsClaude applies style rules directly when generating code/documents
Fallback strategy built-inFont fallback already defined in SKILL.md
Zero dependenciesNo Python, Node.js, or any external tools required
  1. “品牌即代码”模式:将传统存放在 PDF 手册里的品牌规范转化为 AI 可执行的指令——品牌一致性从”人工审查”变为”默认行为”
  2. 精确到值的编码:颜色使用 Hex 值精确指定,而非模糊的”深色”——消除了解释歧义
  3. 智能回退机制:字体规范不仅指定首选字体,还包含完整的跨平台回退链(Poppins → Arial → 系统无衬线)
  4. 层级化的字体策略:根据文本层级(标题 vs 正文)智能区分应用规则,而非粗暴统一
  5. 强调色循环策略:形状颜色循环使用而非随机——视觉上可预期且品牌一致
  6. 极低维护成本:2 个文件,73 行——更新品牌规范只需修改一处

“如果你想为你的企业创建品牌规范 Skill…”

  1. 提取品牌核心元素:收集颜色 Hex 值、字体名称、Logo 使用规则
  2. 定义层级规则:不同文本层级、不同组件类型应用不同的样式规则
  3. 精确编码:所有值使用精确数字/代码,避免模糊描述
  4. 添加回退策略:考虑字体不可用、颜色空间差异等边缘情况
  5. 扩展规则集:根据需求加入间距系统、布局网格、图片风格等

⚠️ 回退字体差异大: Poppins → Arial 的字宽、行高等属性差异显著——标题可能在不同系统上看起来完全不同

⚠️ 跨媒体表现不一致: 屏幕上的品牌颜色在打印时可能看起来不同——需考虑媒体类型

⚠️ 遗忘可访问性: 纯粹的品牌颜色可能不满足 WCAG 对比度要求——需加入无障碍规则

⚠️ 过度约束: 品牌规则应确保一致性而非扼杀创意——留出”品牌安全”的创意空间

⚠️ 颜色适用性有限: Hex 颜色精确但对印刷(CMYK)或特定材料不适用——需要场景判断

  1. “Brand as Code” Pattern: Converting brand guidelines traditionally stored in PDF manuals into AI-executable instructions — brand consistency shifts from “human review” to “default behavior”
  2. Precise Value Encoding: Colors specified with exact Hex values, not vague “dark color” — eliminating interpretation ambiguity
  3. Smart Fallback Mechanism: Font specification includes complete cross-platform fallback chain (Poppins → Arial → system sans-serif)
  4. Hierarchical Font Strategy: Intelligent rule application based on text hierarchy (heading vs body), not blanket uniformity
  5. Accent Color Cycling: Shape colors cycle rather than random — visually predictable and brand-consistent
  6. Minimal Maintenance: 2 files, 73 lines — updating brand guidelines requires one change

“If you want to create a brand guidelines Skill for your enterprise…”

  1. Extract core brand elements: Collect color Hex values, font names, logo usage rules
  2. Define hierarchical rules: Different text levels and component types apply different style rules
  3. Precise encoding: Use exact numbers/code for all values, avoid vague descriptions
  4. Add fallback strategy: Consider edge cases like unavailable fonts, color space differences
  5. Extend rule set: Add spacing systems, layout grids, image styles as needed

⚠️ Fallback font variation: Poppins → Arial has significant differences in character width, line height — headings may look completely different across systems

⚠️ Cross-media inconsistency: Brand colors on screen may look different in print — need to consider media type

⚠️ Accessibility oversight: Pure brand colors may not meet WCAG contrast requirements — need to add accessibility rules

⚠️ Over-constraining: Brand rules should ensure consistency without killing creativity — leave “brand-safe” creative space

⚠️ Limited color applicability: Hex colors are precise but inapplicable to print (CMYK) or specific materials — needs scenario judgment

模式说明适用于...
品牌即代码将品牌规范编码为 AI 可执行的指令集合任何需要品牌一致性的组织
精确值编码所有样式参数使用精确值(Hex、px、pt)颜色、字体大小、间距等需要精确值的场景
层级规则根据文本或组件的层级智能应用样式有层级差异的设计系统(标题/正文/标注)
回退链定义完整的跨平台/跨环境回退路径需要跨系统/浏览器一致性的场景
PatternDescriptionApplies to...
Brand as CodeEncode brand guidelines as AI-executable instruction setAny organization needing brand consistency
Precise Value EncodingUse exact values for all style parameters (Hex, px, pt)Scenarios needing precise values (colors, font sizes, spacing)
Hierarchical RulesSmart style application based on text/component hierarchyDesign systems with hierarchy differences (heading/body/caption)
Fallback ChainDefine complete cross-platform/environment fallback pathsScenarios needing cross-system/browser consistency