Token导航 LogoToken导航

Agent Skill · 效率

abm-brand-voice(abm 品牌之声)

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。

abm-brand-voice 是效率类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为只读;许可证为 MIT-0,第三方扫描已通过,按安装量在效率类的 21,003 个 Skill 里排第 11,931。

总安装 4,351周安装 185GitHub Stars 公开资料未说明下载 1,524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-0
版本
1.0.0

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:abm-brand-voice(abm 品牌之声)
来源仓库:https://github.com/mariokarras/abm-brand-voice
安装命令:
openclaw skills install abm-brand-voice
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。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install abm-brand-voice

这个 Skill 是做什么的

效率类 Skill 把重复的整理、汇总、转换工作交给助手完成,减少手工复制粘贴。

abm-brand-voice 定义与实施品牌声音与风格指南,统一沟通语气与表达方式。适用于营销文案、客服对话与客户沟通场景。提供语气矩阵、词汇表与使用示例。abm-brand-voice 属于效率类 Skill,可作为该场景下的辅助能力补充。需结合行业特性与受众偏好调整风格,避免刻板印象。

建议建立内部审核流程,确保品牌一致性。

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 OpenClaw;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 openclaw skills install abm-brand-voice;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

abm-brand-voice 怎么安装?

在终端执行 openclaw skills install abm-brand-voice。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

abm-brand-voice 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。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。

abm-brand-voice 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

abm-brand-voice 是免费的吗?

来源标注的许可证是 MIT-0。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。

abm-brand-voice 用的人多吗?

公开数据是累计安装 4,351 次,近一周安装 185 次,下载 1,524 次。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。

abm-brand-voice 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

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,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。
  • 标注为只读权限,试用阶段不容易误改本地文件。

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 OpenClaw,想直接扩展它能力的人。
  • 需要处理效率类任务、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。

效率类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
github

GitHub 代码协作 · OpenClaw

7,632,918545需要联网MIT-0通过
nano-pdf

Nano PDF 文档 · OpenClaw

4,249,476231需要联网MIT-0通过
auto-updater

自动更新程序 · OpenClaw

2,964,063386操作浏览器MIT-0通过
video-frames

视频帧 · OpenClaw

2,621,304112需要联网MIT-0通过
gemini

Gemini 助手 · OpenClaw

2,547,39648需要联网MIT-0有提醒
abm-brand-voice

abm 品牌之声 · 本页

4,351—只读MIT-0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

name
brand-voice
description
When the user wants to define, document, or enforce their brand voice and style guide. Use when the user mentions 'brand voice,' 'tone of voice,' 'style guide,' 'voice and tone,' 'writing guidelines,' 'brand personality,' 'tone adaptation,' 'messaging pillars,' 'terminology guide,' 'how should we sound,' or wants consistency across content. Creates .agents/brand-voice-guide.md — a comprehensive style reference that downstream skills use alongside product-marketing-context.md. For foundational positioning context, see product-marketing-context.
metadata
version
1.0.0

Brand Voice

You help users define, document, and enforce a comprehensive brand voice and style guide. This goes deeper than the Brand Voice section in product-marketing-context — it covers voice attributes with examples, tone adaptation by channel and situation, style rules, and terminology governance.

The document is stored at .agents/brand-voice-guide.md.

Workflow

Step 1: Check for Existing Context

First, check if .agents/brand-voice-guide.md already exists.

If it exists:

  • Read it and summarize what's captured
  • Ask which sections they want to update or expand
  • Only gather info for those sections

If it doesn't exist:

Check if .agents/product-marketing-context.md exists. If it does, read the Brand Voice, Customer Language, and Personas sections — use them as a starting point so the user doesn't repeat themselves.

Offer two options:

  1. Build from product-marketing-context (recommended if it exists): Draft V1 from the existing context, then expand each area with the deeper frameworks below
  2. Start from scratch: Walk through each section conversationally

Step 2: Gather Information

Walk through sections conversationally. Don't dump all questions at once — one section at a time, confirm, move on.

Question design — reduce friction, get better answers:

  • Select one from spectrum: For preference fields (formality, authority, emotion), present a scale: "Where do you fall? Formal ← → Casual." Let them pick a point.
  • Select all that apply: For multi-value fields (channels, situations), present a checklist of common options plus "other."
  • Finish this sentence: When users freeze, offer fill-in-the-blank: "Our brand sounds like ___ but never ___."
  • Examples as anchors: Show a concrete example of a good answer before asking. E.g., show a completed voice attribute before asking them to define theirs.
  • Confirm by summarizing: After each section, summarize in 2-3 sentences and ask: "Does this capture it? Anything to adjust?"
  • Stay conversational: These patterns are tools to reach for when the user is stuck, not a rigid form. If the user gives a rich answer unprompted, capture it and move on. Use structured prompts to unblock, not to gate.
  • Match pattern to question type — sections below use → *pattern*: annotations as instructions for you, not text to show the user.

Important: Adapt to offering type and audience. A B2C healthcare service needs different tone guidance than a B2B developer tool. If product-marketing-context exists, use its offering type and audience to shape your questions.


Sections to Capture

Priority guide: Sections 1-4 are the foundation — they give downstream skills enough to produce consistent content. Sections 5-7 add precision and are high-value for teams with multiple writers or agencies. Mark sections as "[to revisit]" if the user runs out of time.

1. Brand Personality

Define the brand as if it were a person.

  • → *finish this sentence*: "If our brand were a person at a dinner party, other guests would describe them as ___."
  • Capture the character sketch as a 2-3 sentence description → *constraining prompt*: "Describe them in 2-3 sentences."
  • Extract 3-5 defining adjectives from the sketch
  • → *finish this sentence*: "Our brand would NEVER be described as ___." — this prevents drift

Output example:

"A knowledgeable colleague who explains complex things simply, celebrates your wins genuinely, and never talks down to you. Not a detached expert, not a hype-driven salesperson."

2. Voice Attributes

Select 3-5 attributes that define how the brand communicates. For each attribute, capture all four dimensions — the "is / is not" pairing prevents the most common brand voice failures.

For each attribute:

  • We are: what this means in practice
  • We are not: the common misinterpretation to avoid
  • This sounds like: an example sentence demonstrating the attribute
  • This does NOT sound like: an example sentence violating it

Ask the user to provide real examples from their existing content if possible. If they don't have content yet, co-write the examples together.

→ *spectrum selection*: Present these spectrums and ask them to pick where they fall on each. This helps them identify their 3-5 attributes:

SpectrumOne EndOther End
FormalityFormal, institutionalCasual, conversational
AuthorityExpert, authoritativePeer-level, collaborative
EmotionWarm, empatheticDirect, matter-of-fact
ComplexityTechnical, preciseSimple, accessible
EnergyBold, energeticCalm, measured
HumorPlayful, wittySerious, earnest

For each attribute they identify, walk through the four dimensions using the example as an anchor before asking.

3. Core Messaging Pillars

3-5 key themes the brand consistently communicates. These are the "big ideas" that should surface across all content.

For each pillar:

  • The theme in one phrase
  • Why it matters to the audience (connect to their pain points or aspirations)
  • How it shows up in content (what kinds of claims, stories, or proof points support it)

→ *constraining prompt*: "If a customer could only remember three things about you, what should those be?"

Rank them by priority → ask: "Which one leads in a headline? Which supports in body copy?"

If product-marketing-context exists, draft pillars from the Differentiation and Proof Points sections and ask the user to validate and rank them.

4. Tone Spectrum

The voice stays consistent, but tone adapts by context. Tone is the emotional inflection applied to the voice.

Key principle to convey: Voice attributes remain fixed. Tone dials them up or down. For example, if a brand is "bold and warm" — in a product launch, dial up boldness; in an incident response, dial up warmth. Neither disappears.

Capture tone by channel → *select all that apply, then fill in tone for each*: "Which of these channels do you use?" Present the table and only fill rows for channels they select:

ChannelTone AdaptationExample
Website/landing pages[their guidance][their example]
Blog/articles
Social (specify platform)
Email marketing
Sales/proposals
Support/help docs
Ads

Capture tone by situation:

SituationTone Adaptation
Product launch / announcement
Incident, outage, or bad news
Customer success / case study
Onboarding / welcome
Price increase or breaking change

Only capture channels and situations that are relevant to the user's business. Don't force a SaaS startup to define press release tone if they've never written one.

5. Style Rules

Specific grammar, formatting, and language rules that ensure consistency across writers. These are the "house rules" — once decided, they shouldn't vary.

Grammar and mechanics:

RuleOptions to present
Oxford commaYes / No
HeadingsSentence case / Title case
ContractionsUse freely / Use sparingly / Avoid
NumbersSpell out 1-9 / Always numerals
Percent"%" / "percent"
Date formatMonth DD, YYYY / DD/MM/YYYY / other
ListsEnd with periods / No periods on fragments

Formatting conventions:

  • Bold and italic usage
  • Link text policy (descriptive vs. "click here")
  • Emoji policy by channel

Punctuation and emphasis:

  • Exclamation mark policy (e.g., max one per paragraph, never in headlines)
  • ALL CAPS policy
  • Ellipsis usage

→ *select one per rule*: Present each grammar rule as a choice: "Oxford comma: (a) Yes (b) No." Help them decide quickly — these are decisions, not debates. If they don't have preferences, suggest sensible defaults and confirm.

6. Terminology

Preferred and avoided terms. This prevents inconsistency and protects positioning.

Preferred terms:

Use ThisNot ThisNotes

Ask for:

  • Product and feature names (official capitalization, full name vs. shorthand)
  • Whether to use "the" before product names
  • Trademark/registration symbols (when required)

Inclusive language guidance:

  • Gender-neutral defaults (they/them for unknown individuals)
  • Person-first language preferences
  • Terms to avoid and their replacements

Jargon management:

  • Terms the audience understands without explanation
  • Jargon that should always be replaced with plain language
  • Acronyms that need spelling out on first use

Competitor and category terms:

  • Preferred category framing (how you name what you do)
  • How to refer to competitors (by name or generically)
  • Terms competitors have coined that you should avoid

If product-marketing-context exists, pull the Glossary and Words to Use/Avoid into this section as a starting point.

7. Audience-Specific Voice Notes

If the brand speaks to multiple distinct audiences (captured in Personas from product-marketing-context), document how voice adapts for each.

For each audience:

  • What they care about (pull from personas)
  • Expertise level (what you can assume they know)
  • How they expect to be addressed
  • Key voice adjustments (more technical? more empathetic? more formal?)

Example: A home healthcare service might speak warmly and reassuringly to family decision-makers, clinically and efficiently to referring physicians, and clearly and patiently to elderly patients.


Step 3: Create the Document

After gathering information, create .agents/brand-voice-guide.md with this structure:

# Brand Voice Guide

*Last updated: [date]*

## Brand Personality
**Character sketch:**
> [2-3 sentence description]

**Defining traits:** [3-5 adjectives]
**We are NOT:** [what the brand is not]

## Voice Attributes

### [Attribute 1]
- **We are**: [what this means]
- **We are not**: [misinterpretation to avoid]
- **Sounds like**: "[example]"
- **Does NOT sound like**: "[counter-example]"

### [Attribute 2]
[same format]

### [Attribute 3]
[same format]

## Core Messaging Pillars

| Priority | Pillar | Why it matters | How it shows up |
|----------|--------|---------------|-----------------|
| 1 | | | |
| 2 | | | |
| 3 | | | |

## Tone Spectrum

**Principle:** Voice stays constant. Tone dials up or down by context.

### By Channel
| Channel | Tone | Example |
|---------|------|---------|
| | | |

### By Situation
| Situation | Tone |
|-----------|------|
| | |

## Style Rules

### Grammar & Mechanics
| Rule | Our choice |
|------|-----------|
| Oxford comma | |
| Headings | |
| Contractions | |
| Numbers | |
| Percent | |
| Date format | |
| Lists | |

### Formatting
- **Bold**: [when to use]
- **Italic**: [when to use]
- **Links**: [policy]
- **Emoji**: [policy by channel]

### Punctuation
- **Exclamation marks**: [policy]
- **ALL CAPS**: [policy]
- **Ellipsis**: [policy]

## Terminology

### Preferred Terms
| Use This | Not This | Notes |
|----------|----------|-------|
| | | |

### Product Names
| Name | Capitalization | Shorthand | Notes |
|------|---------------|-----------|-------|
| | | | |

### Inclusive Language
[guidelines]

### Jargon Guide
| Term | Audience knows it? | Action |
|------|-------------------|--------|
| | Yes / No | Use freely / Define / Replace with... |

### Competitor & Category Terms
**Our category:** [preferred framing]
**Refer to competitors:** [by name / generically]
**Terms to avoid:** [competitor-coined terms]

## Audience-Specific Notes

### [Audience 1]
- **Expertise**: [level]
- **Expects**: [how they want to be addressed]
- **Adjust**: [voice adjustments]

### [Audience 2]
[same format]

Step 4: Confirm and Save

  • Show the completed document
  • Ask if anything needs adjustment
  • Save to .agents/brand-voice-guide.md
  • Tell them: "Content skills will now use this guide alongside your product marketing context. Run /brand-voice anytime to update it."

Tips

  • Start with personality: The character sketch anchors everything else. If the user can describe the brand as a person, the attributes and tone follow naturally.
  • Push for counter-examples: "This does NOT sound like" is often more useful than "this sounds like" — it prevents the most common drift.
  • Don't over-specify early: A startup with one writer doesn't need 30 style rules. Capture what matters now and mark the rest "[to revisit]."
  • Pull from existing content: If they have content they're proud of, use it as voice examples. If they have content they hate, use it as counter-examples.
  • Respect product-marketing-context: Don't re-ask what's already captured there. Build on it.
  • Style rules are decisions, not debates: For grammar choices (Oxford comma, contractions, etc.), present options, help them choose, and move on. These aren't strategic — they just need to be consistent.

平台分布

OpenClaw

80.98%
按下载量换算1,234

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills