Token导航 LogoToken导航

Agent Skill · 研究检索

brand-voice-development(品牌声音发展)

用于辅助音频、音乐、语音转写、语音合成或声音素材处理。它适合让 Agent 生成配乐说明、整理音频流程、调用语音工具或处理播客和视频配音素材。使用时需要确认输入音频来源、输出格式、时长和模型限制;涉及人声克隆、版权音乐或公开发布时,应先核对授权和合规边界。

brand-voice-development 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证为 MIT,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 41,814。

总安装 1,137周安装 46GitHub Stars 16下载 357核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:brand-voice-development(品牌声音发展)
来源仓库:https://github.com/mike-coulbourn/claude-vibes
仓库路径:skills/brand-voice-development
安装命令:
npx skills add https://github.com/mike-coulbourn/claude-vibes --skill brand-voice-development
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/mike-coulbourn/claude-vibes --skill brand-voice-development

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

brand-voice-development 基于主流语音战略方法论,构建可适应不同场景的品牌声音体系。适用于区分 voice(恒定人格)与 tone(情境变化),指导内容调性灵活调整。提供 Mailchimp 等成功案例参考,强调持续练习与反馈迭代的重要性。

使用时应明确核心 voice 特质,再据此设计 tone flex 策略矩阵。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/brand-voice-development

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/mike-coulbourn/claude-vibes --skill brand-voic…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

brand-voice-development 怎么安装?

在终端执行 npx skills add https://github.com/mike-coulbourn/claude-vibes --skill brand-voice-developm…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

brand-voice-development 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

brand-voice-development 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

brand-voice-development 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

brand-voice-development 用的人多吗?

公开数据是累计安装 1,137 次,近一周安装 46 次,下载 357 次,GitHub 16 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

brand-voice-development 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
brand-voice-development

品牌声音发展 · 本页

1,13716需要联网MIT通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Brand Voice Development Framework

Quick reference for developing brand voice and verbal identity using established methodologies from leading voice strategists.

"You have the same voice all the time, but your tone changes. You might use one tone when you're out to dinner with your closest friends, and a different tone when you're in a meeting with your boss." — Mailchimp Content Style Guide

Key Distinction: Voice vs. Tone

ConceptDefinitionConsistency
VoiceBrand's personality expressed through languageAlways consistent
ToneHow voice adapts to different contextsVaries by situation

Analogy: You have the same voice whether talking to your boss or your friends — but your tone changes.


Core Methodologies

1. Bloomstein's Message Architecture & BrandSort

Created by: Margot Bloomstein

The Process:

  1. Sort attribute cards into three piles:

- "Who you are" (current perception) - "Who you'd like to be" (aspirational) - "Who you're NOT" (actively rejected)

  1. Prioritize: From "who you'd like to be," create hierarchy of 8-12 attributes
  2. Unpack abstract terms: Make "innovative" or "trustworthy" concrete
  3. Content audit: Evaluate existing content against the architecture

Key Principle: Balance user needs with brand identity — being exclusively user-focused leads to bland sameness.


2. Nielsen Norman Group's Four Dimensions of Tone

Research-backed framework for plotting voice position:

DimensionSpectrumExample
FormalityFormal ←→ Casual"We appreciate your inquiry" vs "Hey, good question!"
HumorSerious ←→ FunnyStraightforward vs witty
ReverenceRespectful ←→ IrreverentTraditional vs rule-breaking
EnergyMatter-of-fact ←→ EnthusiasticNeutral vs excited

Application:

  • Plot intended position on each dimension
  • A brand CAN be serious AND casual (like a friendly financial advisor)
  • Avoid landing exactly in the middle on all dimensions — that's generic

Research Finding: Casual, conversational tone increased trustworthiness by 0.3 points on a 5-point scale.


3. Aaker's Brand Personality Dimensions

The most influential academic framework (Stanford, 1997):

DimensionSub-Traits
SincerityDown-to-earth, honest, wholesome, cheerful
ExcitementDaring, spirited, imaginative, up-to-date
CompetenceReliable, intelligent, successful
SophisticationUpper class, charming, glamorous
RuggednessOutdoorsy, tough, masculine

Usage: Identify 1-2 primary dimensions for your brand as starting point.


4. "This But Not That" Technique

Used by Slack to add nuance and prevent misinterpretation:

Voice AttributeConstraintWhy It Matters
Confidentbut not cockySets boundary on confidence
Wittybut not sillyHumor with purpose
Friendlybut not ingratiatingWarmth without being desperate
Conversationalbut always appropriateCasual yet professional
Intelligentand treats users as intelligentRespects the audience

Key Principle: The "not that" constraint prevents common misinterpretations.


Complete Verbal Identity System

A comprehensive verbal identity includes 8 components:

#ComponentDescription
1Voice Attributes3-5 core traits with definitions and "this but not that" refinements
2Tone GuidelinesHow voice adapts across contexts, channels, emotional states
3Messaging ArchitecturePurpose, value propositions, elevator pitch, key messages
4Naming ConventionsProduct/feature naming system, hierarchy
5Verbal Distinctive AssetsOwnable phrases, signature words, coined terms
6Vocabulary GuidelinesWords to use/avoid, jargon policy, inclusive language
7Grammar & MechanicsSentence structure, punctuation, contractions, emoji policy
8Examples LibraryBefore/after rewrites, good/bad examples, templates

Tone Adaptation Framework

Tone Matrix by Channel

ChannelTone Adjustment
Social MediaMore casual, engaging, playful
Marketing/AdsConfident, benefit-focused
Customer SupportEmpathetic, helpful, clear
Legal/ComplianceMore formal, precise
Product UIConcise, action-oriented

Tone by Emotional State (Mailchimp's Approach)

Reader's StateTone Response
Happy/successfulMatch their energy, celebrate
ConfusedBe patient, clear, helpful
Frustrated/angryBe empathetic first, then helpful
AnxiousBe reassuring, concrete, specific

Slack's 5 Copy Principles

  1. Be clear and simple — Aim for comprehension
  2. Anticipate (and answer!) the readers' questions — Be relevant
  3. Be intentionally playful and bold — Delight without distracting
  4. Build appropriate emotional connection — Be generous with warmth
  5. Help people envision the possibilities of a better future

The best copy demonstrates at least 3 of these 5 principles.


Voice vs. Messaging vs. Positioning

These are often confused. Clear distinctions:

ElementDefinitionAnswersScope
PositioningStrategic foundation for perception"Why choose us?"Internal strategy
MessagingContent and narratives communicating positioning"What do we say?"External content
VoicePersonality and style in which you communicate"How do we say it?"Applied everywhere

How They Work Together:

  • Positioning determines WHAT makes you different
  • Messaging translates that into WHAT you say
  • Voice determines HOW you say everything

Common Mistakes & Anti-Patterns

#MistakePatternFix
1Being Generic"Innovative, customer-focused, passionate"Use the "airport test" — recognizable without logo?
2InconsistencyDifferent teams, different stylesCentral guidelines, train all teams, voice champions
3Stripping PersonalityProduct specs overwhelming messageVoice makes complexity simple
4Copying TrendsEveryone trying to sound like OatlyGround in authentic brand values
5Ignoring AudienceVoice reflects company, not customerResearch audience preferences first
6Vague Guidelines"Be friendly" without definitionAdd concrete dos/don'ts with examples
7Create Then ForgetNo enforcement or monitoringRegular audits, ongoing training, metrics
8Ignoring AI EraGuidelines don't work with AICreate AI-specific guidelines with prompts

Workshop Exercises

Exercise 1: Adjective Card Sort (15 min)

Sort attributes into "You Are," "You Are Torn," "You Are Not." The "Not" pile is as valuable as "Are."

Exercise 2: Sliding Scale Spectrums (10 min)

Place brand on spectrums: Conservative ←→ Extravagant, Fun ←→ Serious, Traditional ←→ Modern.

Exercise 3: Brand as a Person (15 min)

Questions: What would they do for work? Wear? Never say? Who would they befriend? What annoys them?

Exercise 4: Content Audit (20 min)

Review 10 recent pieces. Read aloud. What feels authentic? What makes you wince?

Exercise 5: Competitive Voice Analysis (20 min)

Review competitors' communications. Identify patterns. Find differentiation opportunities.


Brand Voice in the AI Era

The Challenge

AI produces content at scale, but without clear voice instructions, output sounds generic.

Solutions

AI-Specific Guidelines:

  • Distill voice into 3-5 adjectives with explanations
  • Provide extensive examples
  • Include prompts that work with your voice

Custom GPT Instructions:

  • Build assistants with pre-set voice parameters
  • Upload documentation, samples, approved phrasing
  • Test outputs and refine

Human Oversight:

  • AI doesn't understand brand history or nuance
  • Always review before publishing
  • Use AI for drafts, humans for refinement

Key Statistics

  • Tone impacts trust: Casual tone increased trustworthiness by 0.3 points (Nielsen Norman Group)
  • Voice drives conversion: Monzo achieved 500% increase in overdraft uptake through messaging
  • Voice affects downloads: Monzo's optimized copy led to 15% jump in app download clicks

Exemplar Brands

BrandVoice ApproachWhat to Learn
MailchimpGold-standard style guideComprehensive documentation, open-source
Slack"This but not that" + 5 PrinciplesNuanced constraints, developer-friendly
MonzoTraining across 700+ employeesVoice as business driver, metrics
OatlyIrreverent stream-of-consciousnessDistinctive voice creates recognition

Selection Validation Tests

TestQuestionPass Criteria
Airport TestWould you recognize this without seeing the logo?Immediately identifiable
Consistency TestCan we express this across all touchpoints?Translates everywhere
Audience TestDoes this resonate with our customers?Matches their expectations
Authenticity TestDoes this feel true to the founder/brand?Not forced or performative
Differentiation TestDoes this stand out from competitors?Distinctly different

Templates

See reference/templates.md for:

  • Voice Attribute Card Template
  • Tone Matrix Template
  • Voice Guidelines Document Template
  • Before/After Examples Template
  • AI Voice Prompt Template
  • Voice Summary Card Template
  • Output Validation Checklist

When to Apply This Knowledge

During Voice Discovery

  • Use BrandSort methodology
  • Plot on NNGroup four dimensions
  • Apply Aaker personality dimensions

During Voice Refinement

  • Apply "this but not that" technique
  • Create tone adaptation matrices
  • Develop vocabulary guidelines

During Voice Documentation

  • Complete all 8 verbal identity components
  • Build comprehensive examples library
  • Create AI-ready guidelines

During Validation

  • Run all 5 validation tests
  • Check against 8 common mistakes
  • Verify with airport test

平台分布

OpenCode

25.27%
按下载量换算90

Claude Code

24.71%
按下载量换算88

Gemini CLI

17.71%
按下载量换算63

Antigravity

11.67%
按下载量换算42

Codex

8.3%
按下载量换算30

windsurf

3.09%
按下载量换算11

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

继续浏览同类 Skills