marketing-orchestrator-skill

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Marketing Orchestrator Skill

营销编排器Skill

Skill by ara.so — Marketing Skills collection

由 ara.so 开发的Skill — 营销技能合集

What This Project Does

项目功能

The Marketing Orchestrator is a Claude skill that routes marketing requests to six specialist sub-skills: Brand Strategy & ICP, Ad Copy & Paid Media, Content Creation, Repurpose, Campaign Mode, and GitHub Deploy. It includes:

Automatic routing based on user intent
Client memory that persists across sessions
Self-evaluation of outputs before delivery
Revision loop that improves the skill over time from logged usage patterns

营销编排器是一款Claude Skill，可将营销请求路由至6个专业子技能：品牌策略与ICP、广告文案与付费媒体、内容创作、内容复用、营销活动模式以及GitHub部署。它包含以下特性：

自动路由：基于用户意图进行请求分发
客户记忆：跨会话保留客户信息
自我评估：交付前对输出内容进行自评
修订循环：通过记录使用模式随时间优化技能表现

Installation

安装步骤

This is a Claude AI skill designed for Claude.ai with Skills feature enabled.

Clone the repository:

bash

git clone https://github.com/bryanengel72/marketing-orchestrator.git
cd marketing-orchestrator

Install via Claude Skills interface by uploading the project folder
Verify installation by checking for these files:

marketing-orchestrator/
├── SKILL.md                          # Main orchestrator
├── references/
│   ├── revision-log.md               # Session tracking
│   └── client-profiles/              # Client data
└── sub-skills/                       # Six specialist routes
    ├── brand-strategy-icp/SKILL.md
    ├── ad-copy-paid-media/SKILL.md
    ├── content-creation/SKILL.md
    ├── repurpose/SKILL.md
    ├── campaign-mode/SKILL.md
    └── github-deploy/SKILL.md

这是一款为启用了Skills功能的 Claude.ai 设计的Claude AI Skill。

克隆仓库：

bash

git clone https://github.com/bryanengel72/marketing-orchestrator.git
cd marketing-orchestrator

通过Claude Skills界面上传项目文件夹完成安装
检查以下文件以验证安装成功：

marketing-orchestrator/
├── SKILL.md                          # 主编排器
├── references/
│   ├── revision-log.md               # 会话追踪
│   └── client-profiles/              # 客户数据
└── sub-skills/                       # 6个专业分支
    ├── brand-strategy-icp/SKILL.md
    ├── ad-copy-paid-media/SKILL.md
    ├── content-creation/SKILL.md
    ├── repurpose/SKILL.md
    ├── campaign-mode/SKILL.md
    └── github-deploy/SKILL.md

Project Structure

项目结构

Core Components

核心组件

SKILL.md (Main Orchestrator)

Handles intake and routing
Manages client profile lookup/creation
Performs self-evaluation before delivery
Tracks revision loop (every 3rd session)

Sub-Skills (Six Specialists) Each sub-skill has its own

SKILL.md

with specialist instructions:

```
brand-strategy-icp/
```
— ICP profiles, positioning, brand voice
```
ad-copy-paid-media/
```
— Platform ads with character limits
```
content-creation/
```
— Blogs, social posts, email sequences
```
repurpose/
```
— Multi-platform content adaptation
```
campaign-mode/
```
— Full campaign architecture
```
github-deploy/
```
— Git setup for non-developers

References

```
references/revision-log.md
```
— Session log for improvement tracking
```
references/client-profiles/
```
— Persistent client context

SKILL.md（主编排器）

处理请求接收与路由
管理客户档案的查询与创建
交付前执行自我评估
追踪修订循环（每3次会话触发一次）

子技能（6个专业分支）每个子技能都有独立的

SKILL.md

文件，包含专业操作说明：

```
brand-strategy-icp/
```
— ICP档案、定位、品牌调性
```
ad-copy-paid-media/
```
— 符合字符限制的平台广告
```
content-creation/
```
— 博客、社交帖子、邮件序列
```
repurpose/
```
— 跨平台内容适配
```
campaign-mode/
```
— 完整营销活动架构
```
github-deploy/
```
— 面向非开发者的Git设置指导

参考文件

```
references/revision-log.md
```
— 用于优化追踪的会话日志
```
references/client-profiles/
```
— 持久化客户上下文信息

How to Use

使用方法

Starting a Session

启动会话

The orchestrator automatically handles routing. User says:

"I need Google Ads copy for a new SaaS product"

Orchestrator:

Checks for existing client profile in

references/client-profiles/{client-name}.md

If new, asks intake questions (company name, industry, target audience, etc.)
Routes to
```
ad-copy-paid-media
```
sub-skill
Sub-skill generates output
Self-evaluates against rubric (audience fit, clarity, voice, format, goal alignment)
Logs session to
```
references/revision-log.md
```
Delivers final output with quality check signal

编排器会自动处理路由。用户只需输入：

"我需要为一款新SaaS产品创作Google Ads文案"

编排器执行流程：

检查

references/client-profiles/{client-name}.md

中是否存在现有客户档案

如果是新客户，询问入门问题（公司名称、行业、目标受众等）
将请求路由至
```
ad-copy-paid-media
```
子技能
子技能生成输出内容
根据评估标准（受众适配性、清晰度、调性、格式、目标对齐度）进行自我评估
将会话记录至
```
references/revision-log.md
```
交付带有质量检查标识的最终输出

Client Profile Format

客户档案格式

Client profiles are saved as:

references/client-profiles/acme-corp.md

Example profile structure:

markdown

undefined

客户档案保存为：

references/client-profiles/acme-corp.md

档案示例结构：

markdown

undefined

Acme Corp

Industry: B2B SaaS Target Audience: Mid-market operations managers Brand Voice: Professional, data-driven, empathetic Key Pain Points: Manual workflows, data silos Unique Value Prop: All-in-one automation without IT dependency

行业： B2B SaaS 目标受众： 中端市场运营经理 品牌调性： 专业、数据驱动、共情 核心痛点： 手动工作流、数据孤岛 独特价值主张： 无需IT依赖的一体化自动化解决方案

Past Projects

过往项目

2026-04-15: Google Ads campaign (5 ad variants)
2026-05-02: Blog post series (productivity tips)

2026-04-15：Google Ads营销活动（5种广告变体）
2026-05-02：博客文章系列（生产力技巧）

Notes

注意事项

Avoid technical jargon
Emphasize ROI and time savings

undefined

避免技术术语
强调投资回报率与时间节省

undefined

Revision Loop

修订循环

Every third session, the orchestrator prompts:

"📊 Revision reminder: It's been 3 sessions. Would you like me to analyze 
the revision log and suggest an improvement?"

If user accepts:

Reads
```
references/revision-log.md
```
Identifies most frequent pattern (e.g., "intake questions too long")
Proposes specific fix
Updates relevant sub-skill after approval
Logs change with version bump

每完成3次会话，编排器会提示：

"📊 修订提醒：已完成3次会话。是否需要我分析修订日志并提出优化建议？"

如果用户同意：

读取
```
references/revision-log.md
```
识别最常见的模式（例如："入门问题过长"）
提出具体修复方案
获得批准后更新相关子技能
记录变更并更新版本号

Revision Log Format

修订日志格式

markdown

undefined

markdown

undefined

Revision Log

修订日志

Version 1.2.0 (2026-05-17)

版本1.2.0（2026-05-17）

Reduced brand-strategy intake from 8 to 5 questions
Added auto-skip for returning clients with profiles
Source: Sessions #12, #14, #16 flagged intake length

将品牌策略入门问题从8个减少至5个
为已有档案的回头客添加自动跳过入门问题功能
来源：会话#12、#14、#16标记入门流程过长

Session #16 (2026-05-17)

会话#16（2026-05-17）

Route: brand-strategy-icp Client: New (TechFlow Inc) Quality Check: ✅ Passed all rubric categories Flag: Intake took 4 back-and-forth exchanges

路由： brand-strategy-icp 客户： 新客户（TechFlow Inc） 质量检查： ✅ 通过所有评估类别 标记： 入门流程经过4次来回沟通

Session #15 (2026-05-14)

会话#15（2026-05-14）

Route: ad-copy-paid-media Client: Returning (Acme Corp) Quality Check: ✅ Passed all rubric categories Note: Profile auto-loaded, no redundant questions

undefined

路由： ad-copy-paid-media 客户： 回头客（Acme Corp） 质量检查： ✅ 通过所有评估类别 备注： 档案自动加载，无冗余问题

undefined

Key Workflows

关键工作流

1. Brand Strategy with Document Export

1. 带文档导出的品牌策略

User request:

"Create an ICP profile and positioning doc for my coaching business"

Sub-skill output includes:

Structured ICP with demographics, psychographics, pain points
Positioning framework (category, audience, problem, solution, differentiation)
Brand voice guide with tone attributes and examples
Downloadable Word doc with formatted sections

The sub-skill automatically structures content for

.docx

export via Claude's artifact system.

用户请求：

"为我的教练业务创建ICP档案与定位文档"

子技能输出包含：

结构化ICP，包含人口统计、心理统计、痛点
定位框架（品类、受众、问题、解决方案、差异化）
带有调性属性与示例的品牌调性指南
可下载Word文档，包含格式化章节

子技能会自动调整内容格式，以便通过Claude的工件系统导出为

.docx

文件。

2. Ad Copy with Platform Constraints

2. 符合平台限制的广告文案

User request:

"Write Meta Ads for a webinar registration campaign"

Output includes:

markdown

undefined

用户请求：

"为网络研讨会注册活动创作Meta Ads"

输出内容包含：

markdown

undefined

Ad Variant A

广告变体A

Primary Text (125 chars): Join 500+ marketers mastering AI workflows. Free webinar May 30. Save your seat →

Headline (40 chars): Master AI Marketing in 60 Minutes

Description (30 chars): Live Q&A included. No replay.

Character Counts:

Primary: 125/125 ✅
Headline: 32/40 ✅
Description: 29/30 ✅


Platform limits are enforced during generation.

主文本（125字符）： 加入500+掌握AI工作流的营销人员。5月30日免费网络研讨会。立即预留席位 →

标题（40字符）： 60分钟掌握AI营销

描述（30字符）： 包含直播问答，无回放。

字符计数：

主文本：125/125 ✅
标题：32/40 ✅
描述：29/30 ✅


生成过程中会严格遵守平台限制。

3. Multi-Platform Repurposing

3. 跨平台内容复用

User provides source content (e.g., blog post) and target platforms:

"Repurpose this blog post for LinkedIn, Twitter, and Instagram"

Sub-skill generates:

LinkedIn: Professional narrative (1200-1500 chars), hashtags, engagement hook
Twitter: Thread format (280 chars/tweet), optimized for retweets
Instagram: Visual-first caption, story-friendly formatting

Each is rewritten natively for platform, not truncated.

用户提供源内容（例如博客文章）与目标平台：

"将这篇博客文章改编为LinkedIn、Twitter和Instagram内容"

子技能生成：

LinkedIn： 专业叙事（1200-1500字符）、话题标签、互动钩子
Twitter： 线程格式（每条推文280字符），优化转发率
Instagram： 视觉优先的文案，适配故事格式

所有内容均为原生改写，而非简单截断。

4. Full Campaign Mode

4. 完整营销活动模式

User request:

"Plan a product launch campaign for a new app feature"

Campaign Mode sub-skill:

Defines campaign architecture (awareness → consideration → conversion)
Lists all deliverables in funnel sequence
Creates each asset with connecting narrative thread
Generates launch checklist with dates and owners

Example output structure:

markdown

undefined

用户请求：

"为一款新应用功能规划产品发布营销活动"

营销活动模式子技能：

定义活动架构（认知 → 考虑 → 转化）
按漏斗序列列出所有交付物
创建带有连贯叙事线的每个资产
生成包含日期与负责人的发布清单

示例输出结构：

markdown

undefined

Campaign Architecture

活动架构

Goal: 500 beta signups in 2 weeks Funnel: Teaser posts → Feature explainer → Signup CTA

目标： 2周内获得500个beta注册 漏斗： 预热帖子 → 功能讲解 → 注册呼吁

Deliverables (in sequence)

交付物（按顺序）

Teaser email (1 week before)
LinkedIn announcement post (launch day)
Demo video script (launch day)
Retargeting ad copy (days 3-7)
Follow-up email (1 week after)

[Full copy for each deliverable follows]

预热邮件（发布前1周）
LinkedIn公告帖子（发布当日）
演示视频脚本（发布当日）
重定向广告文案（第3-7天）
跟进邮件（发布后1周）

[以下为每个交付物的完整文案]

Launch Checklist

发布清单

Load email into ESP (Day -1)
Schedule LinkedIn post (Day 0, 9am)
Activate retargeting ads (Day 3)

undefined

将邮件加载至ESP（发布前1天）
安排LinkedIn帖子发布（发布当日9点）
激活重定向广告（第3天）

undefined

5. GitHub Deploy for Non-Developers

5. 面向非开发者的GitHub部署

User request:

"Help me set up GitHub for my marketing files"

Sub-skill provides:

bash

undefined

用户请求：

"帮我为营销文件搭建GitHub"

子技能提供指导：

bash

undefined

Step 1: Create GitHub account at github.com

步骤1：在github.com创建GitHub账户

Step 2: Install Git

步骤2：安装Git

macOS:

brew install git

Windows:

Download from git-scm.com

从git-scm.com下载

Step 3: Configure Git

步骤3：配置Git

git config --global user.name "Your Name" git config --global user.email "your.email@example.com"

Step 4: Create repository on GitHub (via web UI)

步骤4：在GitHub上创建仓库（通过网页界面）

Repository name: marketing-files

仓库名称：marketing-files

Visibility: Private

可见性：私有

Step 5: Initialize local folder

步骤5：初始化本地文件夹

cd ~/Documents/marketing-files git init git remote add origin https://github.com/YOUR_USERNAME/marketing-files.git

Step 6: First commit

步骤6：首次提交

git add . git commit -m "Initial commit" git push -u origin main

Daily workflow:

日常工作流:

git add . git commit -m "Updated campaign docs" git push


Includes troubleshooting for common errors (authentication, branch naming, merge conflicts).

git add . git commit -m "Updated campaign docs" git push


包含针对常见错误（认证、分支命名、合并冲突）的故障排除方案。

Self-Evaluation Rubric

自我评估标准

Before delivering output, the orchestrator evaluates against:

Audience Fit: Does this match the client's target audience profile?
Clarity: Is the language clear and jargon-appropriate?
Voice Consistency: Does it match the client's brand voice?
Platform Format: Does it meet platform-specific requirements (char limits, structure)?
Goal Alignment: Does it directly support the stated objective?

Outputs are silently revised if any category fails. Final delivery includes:

✅ Quality check: Passed all rubric categories

交付输出前，编排器会根据以下标准进行评估：

受众适配性： 是否匹配客户的目标受众档案？
清晰度： 语言是否清晰，术语使用是否恰当？
调性一致性： 是否符合客户的品牌调性？
平台格式： 是否满足平台特定要求（字符限制、结构）？
目标对齐度： 是否直接支持既定目标？

如果任何类别不达标，输出内容会被自动修订。最终交付时会包含：

✅ 质量检查：通过所有评估类别

Configuration

配置

Custom Sub-Skill

自定义子技能

To add a new specialist route:

Create folder:
```
sub-skills/new-specialist/
```
Add
```
SKILL.md
```
with specialist instructions:

markdown

undefined

添加新的专业分支：

创建文件夹：
```
sub-skills/new-specialist/
```
添加包含专业操作说明的
```
SKILL.md
```
：

markdown

undefined

New Specialist Sub-Skill

新专业子技能

Scope

范围

Handles [specific marketing task category]

处理[特定营销任务类别]

Intake Questions

入门问题

Question 1
Question 2

问题1
问题2

Output Format

输出格式

[Structured template]

[结构化模板]

Self-Evaluation Focus

自我评估重点

[Task-specific quality criteria]


3. Update main `SKILL.md` routing logic to include new route trigger phrases

[任务特定质量标准]


3. 更新主`SKILL.md`中的路由逻辑，添加新分支的触发短语

Environment Variables

环境变量

For API integrations (not included in base skill):

bash

export OPENAI_API_KEY=your_key_here
export ANTHROPIC_API_KEY=your_key_here

用于API集成（基础技能未包含）：

bash

export OPENAI_API_KEY=your_key_here
export ANTHROPIC_API_KEY=your_key_here

Troubleshooting

故障排除

Client Profile Not Loading

客户档案未加载

Symptom: Orchestrator asks intake questions for returning client

Fix:

Check filename matches client name (kebab-case):
```
acme-corp.md
```
Verify file location:
```
references/client-profiles/acme-corp.md
```
Confirm markdown syntax is valid

症状： 编排器对回头客仍询问入门问题

修复方案：

检查文件名是否与客户名称匹配（短横线分隔格式）：
```
acme-corp.md
```
验证文件位置：
```
references/client-profiles/acme-corp.md
```
确认Markdown语法有效

Revision Log Not Tracking

修订日志未追踪

Symptom: No entries appear in

references/revision-log.md

Fix:

Ensure file exists at
```
references/revision-log.md
```
Check file permissions (must be writable)
Verify orchestrator is logging at end of each session

症状：

references/revision-log.md

中无条目

修复方案：

确保文件存在于
```
references/revision-log.md
```
检查文件权限（必须可写）
验证编排器在每次会话结束时是否记录日志

Sub-Skill Not Routing

子技能路由错误

Symptom: Wrong specialist handles request

Fix:

Check user request includes clear trigger words (e.g., "ad copy", "campaign", "repurpose")
Review routing logic in main
```
SKILL.md
```
If ambiguous, orchestrator should ask clarifying question before routing

症状： 错误的专业分支处理请求

修复方案：

检查用户请求是否包含明确触发词（例如："广告文案"、"营销活动"、"复用"）
查看主
```
SKILL.md
```
中的路由逻辑
如果请求模糊，编排器应在路由前询问澄清问题

Common Patterns

常见模式

Pattern: Resume Previous Session

模式：恢复过往会话

User returns with:

"Continue working on the Acme Corp campaign from last week"

Orchestrator:

Loads
```
references/client-profiles/acme-corp.md
```
Reads "Past Projects" section for context
Confirms campaign details with user
Routes to appropriate sub-skill

用户返回时输入：

"继续处理上周Acme Corp的营销活动"

编排器执行：

加载
```
references/client-profiles/acme-corp.md
```
读取“过往项目”部分获取上下文
与用户确认活动细节
将请求路由至合适的子技能

Pattern: Batch Content Request

模式：批量内容请求

User:

"Create 5 LinkedIn posts about remote work productivity"

Orchestrator routes to

content-creation

, which:

Generates all 5 posts in single session
Ensures variety in hooks and formats
Numbers each post for easy reference
Offers to add to content calendar

用户输入：

"创作5篇关于远程工作生产力的LinkedIn帖子"

编排器将请求路由至

content-creation

子技能，该子技能：

在单次会话中生成全部5篇帖子
确保钩子与格式多样化
为每个帖子编号以便参考
提供添加至内容日历的选项

Pattern: Quick Edit

模式：快速编辑

User:

"Make that last ad copy more concise"

Orchestrator:

Retrieves previous output from session context
Applies edit instruction
Re-runs self-evaluation
Delivers revised version with quality check

用户输入：

"将上次的广告文案改得更简洁"

编排器执行：

从会话上下文中检索之前的输出
应用编辑指令
重新执行自我评估
交付带有质量检查的修订版本

Advanced Usage

进阶用法

Chaining Multiple Routes

多路由链式调用

User:

"Create a brand voice guide, then write 3 ads using that voice"

Orchestrator:

Routes to
```
brand-strategy-icp
```
first
Saves output to client profile under "Brand Voice" section
Routes to
```
ad-copy-paid-media
```
with voice guide context
Generates ads conforming to established voice

用户输入：

"创建品牌调性指南，然后使用该调性创作3条广告"

编排器执行：

先将请求路由至
```
brand-strategy-icp
```
将输出保存至客户档案的“品牌调性”部分
携带调性指南上下文将请求路由至
```
ad-copy-paid-media
```
生成符合既定调性的广告

Export Formats

导出格式

For document-heavy routes (Brand Strategy, Campaign Mode):

Orchestrator structures output for Claude's artifact system
User can download as
```
.docx
```
,
```
.md
```
, or
```
.pdf
```
Formatting includes headers, bullet lists, tables

针对文档密集型分支（品牌策略、营销活动模式）：

编排器会调整输出格式以适配Claude的工件系统
用户可下载为
```
.docx
```
、
```
.md
```
或
```
.pdf
```
格式
格式包含标题、项目符号列表、表格

Testing the Skill

技能测试

Test 1: New Client Intake

测试1：新客户入门

User: "I need a blog post about AI in healthcare"
Expected: Intake questions → content-creation route → quality check

用户："我需要一篇关于医疗AI的博客文章"
预期：询问入门问题 → 路由至content-creation → 质量检查

Test 2: Returning Client

测试2：回头客

User: "Create another ad set for Acme Corp"
Expected: Auto-load profile → ad-copy route → no redundant questions

用户："为Acme Corp创建另一组广告"
预期：自动加载档案 → 路由至ad-copy → 无冗余问题

Test 3: Revision Loop

测试3：修订循环

After 3rd session, expected: "📊 Revision reminder..." prompt
User: "Yes, analyze the log"
Expected: Pattern analysis → proposed fix → update confirmation

完成3次会话后，预期："📊 修订提醒..."提示
用户："是的，分析日志"
预期：模式分析 → 提出修复方案 → 确认更新

License

许可证

MIT License — Free to use, adapt, and extend.

MIT许可证 — 可免费使用、改编与扩展。