10 Claude Skills for Marketing: Copy-Paste SKILL.md Files

Sources:

• HubSpot / Futurepedia: "Guide to Claude Skills" (Futurepedia, June 25, 2026) — 5 skills: Brand Voice, Audience Profile, Hyper-Personalised Content Engine, Meeting Notes Synthesizer, YouTube Launch Kit
• YouTube: "10 Claude Skills for Marketing" (transcript) — 5 additional skills: Editorial Calendar Builder, Cold Email Personaliser, Competitive Brief, LinkedIn Post Writer, Campaign Brief

Claude forgets everything when a conversation ends. You re-explain your brand voice. You re-describe your audience. You remind it what tone you use. Every session from scratch.

Skills fix this. A skill is a permanent instruction set Claude reads at the start of every relevant conversation. You write it once. Every session after that starts already knowing who you are, how you write, who you are writing for, and what the rules are.

This lead mag has 10 production skills for marketing teams, founders, and GTM engineers. All 10 are written as full copy-paste SKILL.md files. No building from scratch. Copy, customise the bracketed fields, install, and run.

How skills are different from prompts:

• Prompts: you write them fresh each session. Claude uses them once.
• Skills: you write them once. Claude loads them automatically across all future sessions whenever the trigger matches.
• Projects: a shared context space. Skills are the instructions inside it.

How skills are different from Projects:

• Projects give Claude a persistent memory of what you are working on.
• Skills tell Claude how to do a specific type of work, consistently, every time that work comes up.
• The combination (Project + Skills) is what serious operators use.

How to install any skill in this lead mag

Method 1: Claude Settings (chat, Cowork, or mobile app)

# Settings > Customize > Skills > Add New Skill
# Paste: Name, Description, Instructions
# Save. Claude loads it automatically from that session forward.

Method 2: Claude Code (SKILL.md file)

# Create the skill folder
mkdir -p .claude/skills/[skill-name]

# Save the full SKILL.md below into that folder
# Then invoke with /[skill-name] or use the trigger phrase

Method 3: Install from catalog

# Settings > Customize > Skills > Browse
# Search for the skill type
# Click Install

The description field is the most important part. It tells Claude when to load the skill automatically. Write it as a specific trigger situation, not a topic.

Wrong: "For marketing work"

Right: "When I am writing LinkedIn posts, social captions, or any content for [BRAND], load the brand voice rules and apply them automatically"

Three mistakes that make skills invisible

From the Futurepedia guide:

1. The description is a topic, not a trigger. "For content creation" describes what you do, not when Claude should act. Write the specific situation: "when I paste a video transcript and ask for social posts".
2. The instructions are too vague. "Write in my brand voice" tells Claude nothing. Your brand voice skill needs specific vocabulary, sentence rhythm, banned phrases, and examples of good and bad copy. Without those specifics, Claude interprets the instruction generically.
3. The skill tries to do too many things. One skill, one job. A skill that covers brand voice, audience profile, content calendar, and email sequence will either fail to trigger or produce inconsistent output. Split it. Brand voice is one skill. Audience profile is another.

Source: Futurepedia / HubSpot guide (June 2026)

The foundation skills. Install Skill 1 and Skill 2 before any of the others. Every content and outreach skill in this lead mag will produce better output once Claude knows your voice and your audience.

Skill 1: Brand Voice

What it does: teaches Claude your tone, sentence style, and hard limits so every piece of content sounds like you, not like a machine. Install this first. Every other content skill builds on it.

---
name: brand-voice
description: >
  When I am writing any content for [BRAND NAME] including LinkedIn posts, emails,
  social captions, campaign copy, or website copy, load this brand voice skill
  automatically and apply it to every output without me asking.
---

# Brand Voice: [BRAND NAME]

## Who we are
[COMPANY NAME]. [One sentence: what you do and who you do it for].

## Tone
WE ARE: [3-5 adjectives: e.g. direct, peer-level, pragmatic, specific, dry]
WE ARE NOT: [3-5 anti-adjectives: e.g. corporate, salesy, preachy, vague, motivational]

## Sentence style
- [Describe rhythm: e.g. mostly short sentences, occasional longer ones for contrast]
- [Describe structure: e.g. one idea per line, max 3 lines per paragraph]
- [Describe energy: e.g. confident but not loud, matter-of-fact not enthusiastic]

## Our audience
[Title and role]. [Their pain]. [What they are trying to achieve].
They are [time-poor / sceptical of hype / technically literate]. Write to that.

## What we never say
- No em dashes anywhere
- No hashtags in post bodies
- No: leverage, synergy, bandwidth, streamline, foster, genuinely, honestly
- No filler openers: I hope this finds you well, I wanted to reach out, Excited to share
- No generic CTAs: Let me know if you are interested, Feel free to reach out
- No motivational closer: You have got this, The world needs your story

## What we always do
- One idea per line
- Three lines maximum per paragraph
- Lead with specifics: a number, a name, a real example
- British English
- Pair every claim with proof: a result, a client name, or a concrete before/after

## LinkedIn specifically
- Hook earns See More in 2 lines
- No hashtags in the body
- One CTA per post
- No questions as hooks unless they are genuinely surprising

## Email specifically
- Under 100 words for cold outreach
- Subject line under 33 characters
- First line references a specific signal, never a compliment
- One CTA, not a list of options

## Examples of our voice (PASTE 2-3 real sentences or paragraphs here)
[EXAMPLE 1]
[EXAMPLE 2]
[EXAMPLE 3]

## Examples of what sounds wrong
[BAD EXAMPLE 1 + note on why]
[BAD EXAMPLE 2 + note on why]

Skill 2: Audience Profile

What it does: gives Claude a deep picture of who it is writing for. Their vocabulary, frustrations, what makes them act, and what makes them stop reading. Without this, Claude writes to a generic professional. With it, Claude writes to a specific person.

---
name: audience-profile
description: >
  When I am writing content, emails, or ads for [BRAND NAME], load this audience
  profile automatically. Use it to calibrate word choice, assumed knowledge level,
  pain language, and what makes this specific audience take action.
---

# Audience Profile: [BRAND NAME]

## Primary persona
Name: [give them a name, e.g. "Agency Annie" or "Founder Faisal"]
Title: [e.g. Head of Growth, Founder, RevOps Director]
Company type: [e.g. B2B SaaS, outbound agency, professional services]
Company size: [e.g. 10-100 employees]
Geography: [e.g. UK, US, EU]

## Their day-to-day reality
[Describe what they actually do all day. Be specific. What tabs are open? What meetings do they hate? What is always on fire?]

## Their top 3 frustrations
1. [Frustration 1 - in their words, not marketing language]
2. [Frustration 2]
3. [Frustration 3]

## Their goals
1. [Goal 1 - what they actually want to achieve, not what we think they want]
2. [Goal 2]

## What makes them stop reading
- [Trigger 1: e.g. anything that sounds like a sales pitch in the first line]
- [Trigger 2: e.g. jargon they do not recognise]
- [Trigger 3: e.g. vague claims without proof]

## What makes them take action
- [Trigger 1: e.g. a specific number that proves the claim]
- [Trigger 2: e.g. a name they recognise using the tool]
- [Trigger 3: e.g. a deadline or a forcing function]

## Their vocabulary
They say: [list the exact words and phrases they use about their work]
They do NOT say: [list the words we might use that they would find alien or annoying]

## What they are sceptical about
[List the specific claims or formats that make this audience roll their eyes. AI-generated content. Overpromising ROI. Generic case studies. Etc.]

## Secondary personas (if applicable)
[PERSONA 2 NAME]: [Title, company type, key difference from primary persona]
[PERSONA 3 NAME]: [Same format]

Sources: Futurepedia guide (Skills 3 and 5) + YouTube transcript (Skill 6 and 7)

Four content production skills: paste a transcript and get a week of social posts, LinkedIn posts written in your voice, a YouTube launch kit, and a full editorial calendar.

Skill 3: Hyper-Personalised Content Engine

What it does: paste a raw video or podcast transcript and get a week of LinkedIn posts, X/Twitter threads, and an SEO article, all in your brand voice. Mines the transcript for the real ideas rather than just summarising it.

---
name: content-engine
description: >
  When I paste a video transcript, podcast transcript, or any long-form source
  content and ask for social posts, LinkedIn posts, threads, or an article,
  load the content engine skill and run the full repurposing workflow.
  Always load the brand-voice skill alongside this one.
---

# Content Engine

## Step 1: Mine the source, do not summarise it
Before writing anything, extract from the transcript:
- The 3 most shareable claims or contrarian takes
- The 1 specific story, result, or example that proves the main point
- The vocabulary the speaker uses (their exact words, not polished paraphrases)
- Any numbers, names, or specific details that could serve as hooks

Do not summarise the content. Mine it. There is a difference.

## Step 2: Produce the week of content

From one transcript, produce:

1. THREE LinkedIn posts
   - Each must use a different hook mechanism (contrarian / specific number / story opener)
   - Each must stand alone: a reader who sees only this post gets full value
   - Under 200 words each
   - Apply brand-voice skill throughout
   - End with one CTA: a question, a comment keyword, or a follow prompt

2. TWO X/Twitter threads
   - Thread 1: list format (5-7 tweets, each one a standalone insight)
   - Thread 2: story format (hook tweet, 4-5 build tweets, closer with CTA)
   - Hook tweet must work standalone if the reader does not click More
   - Under 280 characters per tweet

3. ONE SEO article outline
   - Target a specific search query the content answers (state it explicitly)
   - H1 as a question
   - 4-6 H2s as sub-questions
   - Answer Block (40-60 words answering the H1 directly, goes in first 100 words)
   - Key Takeaways section
   - FAQ (4 questions, edge cases not restatements of H2s)

## Rules
- Never use the same hook mechanism twice in the same content set
- Never invent a stat that was not in the source material
- Never write a generic opener: every first line must earn the second
- Apply brand-voice skill to all outputs
- Flag any section where the source did not give you enough material

Skill 4: LinkedIn Post Writer

What it does: writes a single LinkedIn post in your brand voice using a selected hook formula. Applies the content moat check before returning the draft. More controlled than the content engine when you have one specific idea.

---
name: linkedin-post-writer
description: >
  When I say write a LinkedIn post, draft a post, or give me a LinkedIn post about
  [TOPIC], load this skill. Always load brand-voice and audience-profile alongside it.
---

# LinkedIn Post Writer

## Before writing

Ask me two things if I have not provided them:
1. What is the one idea this post is about? (One sentence)
2. What is the specific proof, example, or number behind it?

If I have provided both, skip the questions and write.

## Hook selection

Select one hook mechanism based on the idea:

| Hook type | Use when |
|---|---|
| Specific number | There is a stat, metric, or count that surprises |
| Contrarian claim | The post challenges something most people believe |
| Story opener | There is a specific moment, failure, or result to lead with |
| ICP direct address | The post is written for a specific title or situation |
| Stakes escalation | The post reveals a consequence most people have not considered |
| Progressive milestone | The post is a before/after or growth narrative |

Name the hook type used at the top of your output so I can review the choice.

## The content moat check

Before returning the draft, check:
1. Is there a narrative, data, or physical moat in this content?
2. Could this exact post have been written by someone who typed the topic into Claude?
3. Is there at least one specific detail (a name, a number, a quote, a date) that only the author could have included?

If question 2 is yes: add one bracketed placeholder [SPECIFIC DETAIL NEEDED] where the author must add their own experience before posting.

## Output format

Return:
- Hook type used
- The post (under 200 words)
- One suggested CTA (comment keyword, question, or follow prompt)
- One suggested A/B hook variant using a different mechanism

## Rules
- No em dashes anywhere
- No hashtags in the body
- One idea per line
- Three lines maximum per paragraph
- Never start with I, We, or the company name
- Apply brand-voice skill throughout

Skill 5: YouTube Launch Kit

What it does: paste a YouTube transcript and get back five title options, a search-friendly description, timestamped chapters pulled from the actual content, tags, thumbnail text ideas, and a pinned comment ready to drive engagement.

---
name: youtube-launch-kit
description: >
  When I paste a YouTube video transcript and ask for titles, description, chapters,
  tags, or anything to do with publishing a YouTube video, load this skill.
---

# YouTube Launch Kit

## Step 1: Analyse the transcript
Before writing anything, identify:
- The core question or promise this video answers
- The most surprising or specific claim in the video
- The key moments where the topic shifts (these become chapters)
- The vocabulary the creator uses (their words, not YouTube-optimised paraphrases)

## Step 2: Produce the full launch kit

### Titles (5 options)
Write 5 title options across these formulas:
1. How to [SPECIFIC OUTCOME] without [COMMON OBSTACLE] (under 60 characters)
2. I [DID SOMETHING] for [TIME PERIOD]. Here is what happened. (curiosity/story)
3. The [SPECIFIC NUMBER] [THING] that [SPECIFIC OUTCOME] ([NUMBER] in [TIMEFRAME]) (proof-led)
4. Why [COMMON BELIEF] is wrong (and what to do instead) (contrarian)
5. [SPECIFIC AUDIENCE]: [SPECIFIC PROMISE] (direct address)

For each title: mark it as SHORT (under 50 chars) or FULL (50-70 chars).

### Description
Structure:
- First 125 characters: the hook (appears before Show More, must earn the click)
- Lines 2-4: what the video covers in plain language (not a summary, a promise)
- 3-5 bullet points of specific things the viewer will learn or get
- Timestamps block (from Step 3 below)
- 1-2 relevant links (if provided)
- CTA: subscribe, comment prompt, or relevant offer

### Chapters (timestamped)
Pull from the actual transcript. Format:
0:00 [Intro or Hook moment]
[TIMESTAMP] [Chapter title - 4 words max, no clickbait]
[TIMESTAMP] [Chapter title]
[TIMESTAMP] [Chapter title]
[TIMESTAMP] [Outro or CTA]

Do not invent timestamps. Use the actual structure of the transcript.

### Tags (10-15)
Mix of: exact match (what the video is literally about), related terms (adjacent topics), and long-tail phrases (specific questions this video answers). No generic one-word tags.

### Thumbnail text (3 options)
Under 5 words each. Must work at thumbnail size (visible from across the room).
Option A: Number-led
Option B: Outcome-led
Option C: Curiosity-led

### Pinned comment
Under 100 words. Includes: one sentence on what the video covers, one soft CTA (subscribe, comment keyword, or question to drive replies). Sounds human, not promotional.

## Rules
- Never invent a timestamp. Pull only from what is in the transcript.
- Titles must be specific. No generic YouTube formulas with empty brackets.
- Description first 125 characters must not start with the channel name.

Skill 6: Editorial Calendar Builder

What it does: takes your content pillars, posting frequency, and channel mix and builds a 4-week editorial calendar. Includes post angles, hook types, formats, and which pillar each post serves. Anti-repetition logic prevents the same mechanism appearing twice in one week.

---
name: editorial-calendar
description: >
  When I ask to plan my content, build a content calendar, schedule posts,
  or plan [N] weeks of LinkedIn content, load this skill.
---

# Editorial Calendar Builder

## Step 1: Intake
Ask me if I have not already provided:
1. Content pillars: what are the 3-5 topics I post about with authority?
2. Posting frequency: how many posts per week per channel?
3. Channels: LinkedIn only / LinkedIn + X / LinkedIn + X + Instagram?
4. Any upcoming events, launches, or topics that must appear in this period?
5. Any formats I want to test this month (carousel, video, lead magnet post)?

## Step 2: Build the calendar

For each post slot, produce:
- Date and channel
- Pillar it serves
- Post angle (the specific idea, not just the topic)
- Hook type (contrarian / number / story / ICP address / stakes / milestone)
- Format (text post / carousel / lead magnet CTA / video script brief)
- Lead magnet opportunity (yes/no, and if yes, what the CTA would be)

Format as a table:
| Date | Channel | Pillar | Angle | Hook type | Format | Lead magnet? |

## Anti-repetition rules
- No hook type appears more than twice in the same week
- No pillar appears more than twice in the same week (unless only one pillar exists)
- No two consecutive posts use the same format
- At least one lead magnet CTA post per two weeks

## Step 3: Return a second output
After the calendar table, return a list of the 3 posts most likely to perform well and why, based on hook strength and pillar demand for this ICP.

## Rules
- Never schedule the same angle twice (even in different formats)
- Apply brand-voice skill to any copy you write within the calendar
- Flag any week that is heavily weighted toward one pillar

Sources: Futurepedia guide (Skill 4) + YouTube transcript (Skills 7, 8, 9)

Three operational skills that handle the work that happens between content: meeting capture, cold outreach, competitive intelligence.

Skill 7: Meeting Notes Synthesizer

What it does: paste a messy meeting transcript and get back a clean recap: decisions made, action items with owners and due dates, open questions flagged, and a one-paragraph summary for anyone who was not in the room.

---
name: meeting-notes
description: >
  When I paste a meeting transcript or say clean up these notes, synthesise this
  call, or write up the meeting, load this skill. Works on any meeting type:
  client calls, internal standups, sales discovery calls, strategy sessions.
---

# Meeting Notes Synthesizer

## Step 1: Classify the meeting type
From the transcript, determine:
- Meeting type: client call / internal / sales discovery / strategy / other
- Attendees mentioned (names and roles if identifiable)
- Duration (if timestamps are visible)

This classification affects the output format.

## Step 2: Extract the four signals

### Decisions made
Only include items where the transcript shows agreement or confirmation. Not discussion or consideration. A decision is something that was agreed, not something that was mentioned.
Format: [Decision] - agreed by [who]

### Action items
Each action item must have:
- What: the specific task
- Who: the person responsible (name or role)
- By when: date or trigger condition (if stated)
Format: [Task] - [Owner] - [Due date or trigger]

If no due date was stated: mark as [No date set] and flag it.

### Open questions
Things raised but not resolved. Include the person who raised it if identifiable.
Format: [Question] - raised by [who] - [status: unresolved / to follow up]

### One-paragraph summary
For anyone who was not in the room. Under 80 words. What the meeting was about, what was decided, what happens next. No jargon.

## Step 3: Sales discovery calls (additional output)
If the meeting is a sales discovery call, also produce:
- Pain in their exact words (not your interpretation)
- Decision process (who else is involved, what does approval look like)
- Timeline (is there a forcing function?)
- What they said they need to see next
- Agreed next step with specific date (if stated)
- ICP fit: hot / warm / cold based on the conversation

## Rules
- Never invent a decision that was not explicitly made in the transcript
- Never invent a due date. If none was stated, flag it.
- Never summarise a discovery call without the pain in their words
- Flag anything that required a decision but did not get one
- Keep the summary under 80 words

Skill 8: Cold Email Personaliser

What it does: takes a prospect's LinkedIn URL or company website and produces a signal-led cold email first line, subject line variants, and a 3-touch sequence. Applies the personalisation waterfall (hiring signal first, then tech stack, then positioning, then fallback) and runs a QA check before returning.

---
name: cold-email-personaliser
description: >
  When I give you a LinkedIn URL, a company name, or a prospect profile and ask
  for a cold email, first line, outreach angle, or email sequence, load this skill.
  Always apply brand-voice skill to all output.
---

# Cold Email Personaliser

## My offer
[PASTE YOUR ONE-SENTENCE OFFER HERE]

## My ICP
[PASTE YOUR ICP DEFINITION HERE: title, company type, size, function]

## Personalisation waterfall
Check signals in this order. Use the first one you can verify:

1. HIRING SIGNAL
   Where: company careers page, LinkedIn jobs
   Use if: they are hiring a role that indicates the pain your offer solves
   Opener format: reference the specific role and the implication for their situation

2. TECH STACK SIGNAL
   Where: BuiltWith data, job posting requirements, their website footer
   Use if: they use a specific tool that indicates a relevant pain
   Opener format: name the tool and connect it to the problem

3. CONTENT / POSITIONING SIGNAL
   Where: their LinkedIn posts, their about page, their website tagline
   Use if: they have said something publicly that connects to what you do
   Opener format: reference what they said and why it resonated

4. COMPANY MILESTONE SIGNAL
   Where: news, LinkedIn company page, Crunchbase
   Use if: funding, new hire, launch, or growth announcement in last 90 days
   Opener format: name the milestone and connect it to the pain or opportunity

5. FALLBACK
   Use something specific from their website or LinkedIn that is still true of them.
   Never use a generic opener. The fallback must still be specific.
   Flag it as FALLBACK so I know.

## Email structure (strict)
- Line 1: personalised opener from the waterfall (under 25 words)
- Line 2: connect their situation to the problem you solve (under 20 words)
- Line 3: one-sentence offer (what you do, for whom, with proof if available)
- Line 4: soft CTA (one question or low-friction ask, not a calendar link)
- Total: under 100 words

## Subject line (3 variants)
- Under 33 characters each (avoids mobile truncation)
- No all-caps, no clickbait, no question marks in all three
- Each must be meaningfully different from the others

## QA checks (run before returning)
- Word count under 100
- No em dashes
- No banned phrases from brand-voice skill
- No fabricated facts about the prospect
- Signal confidence: high (verified) / medium (inferred) / low (assumed)
- Flag anything that should be checked by a human before sending

## 3-touch sequence (if requested)
Touch 1 (Day 1): above format
Touch 2 (Day 3): different angle, different signal if available. Under 80 words.
Touch 3 (Day 7): honest breakup. No guilt. Leaves door open. Under 50 words.
Different subject line per touch.

Skill 9: Competitive Intelligence Brief

What it does: researches a competitor from their website, G2 reviews, job postings, and LinkedIn and returns a structured brief: their current positioning, top customer complaints, what their hiring signals say about their roadmap, where you win, and what to say when their name comes up on a call.

---
name: competitive-brief
description: >
  When I say research [COMPETITOR], brief me on [COMPETITOR], or ask what to say
  when [COMPETITOR] comes up, load this skill. Use web search to pull live data.
---

# Competitive Intelligence Brief

## Sources to check (in order)
1. Their website: homepage, pricing page, about page
2. G2 reviews: last 90 days, filter for most recent. Focus on 1, 2, and 3-star reviews.
3. LinkedIn: their company page posts from last 30 days
4. Job postings: what roles are they hiring? What do the requirements reveal?
5. Crunchbase or news: any funding, hires, or announcements in last 90 days

## Output structure

### Their positioning (in their own words)
[Paste or paraphrase exactly how they describe themselves on their homepage. Their headline. Their subheadline. Do not interpret. Quote directly.]

### Top 5 customer complaints (from G2)
[Each complaint in the customer's own words. Not your interpretation.]

### What their hiring signals say
[Which departments are they growing? What tools are listed in job requirements? What does this suggest about their roadmap or priorities?]

### Where we win
[Specific claims we can make that they cannot make back, based on the G2 gaps and positioning review.]

### What to say when their name comes up on a call
[One paragraph, in our brand voice. Acknowledge them specifically. Name one real weakness without being aggressive. Pivot to the claim we can make that they cannot.]

### The one thing they are doing well
[One honest sentence. Knowing what they do well prevents us from dismissing them to a prospect who has used them.]

## Rules
- Verify any pricing or feature claim before including it. Flag anything you cannot confirm.
- Never invent a G2 complaint. Only use what is in the actual reviews.
- Apply brand-voice skill to the on-call section
- If you cannot access G2 or the website, say so and note what you could not retrieve

Source: YouTube transcript

The final skill: a full campaign brief from a one-line prompt. Takes your offer, ICP, and goal and returns a complete brief that any designer, copywriter, or agent can execute from.

Skill 10: Campaign Brief Generator

What it does: takes a campaign goal and produces a full brief: positioning statement, messaging hierarchy, channel sequence, asset list, success metrics per channel, and the one assumption the campaign lives or dies on. Applies the honest advisor check before returning.

---
name: campaign-brief
description: >
  When I say brief a campaign, plan a campaign, write a brief for [CAMPAIGN TYPE],
  or I need to launch [OFFER/PRODUCT], load this skill.
  Always load brand-voice and audience-profile alongside it.
---

# Campaign Brief Generator

## Step 1: Intake
Ask me if I have not already provided:
1. What is this campaign for? (product launch / lead magnet / event / offer / content series)
2. Who is the primary audience? (use audience-profile skill or describe)
3. What is the single desired action? (sign up / book a call / download / purchase)
4. What is the timeline? (start date, end date or duration)
5. What channels are available? (LinkedIn / email / paid / organic / all)
6. What proof or case study can we use? (client name, result, or story)

## Step 2: The honest advisor check
Before writing the brief, name:
- The one assumption this campaign lives or dies on
- The cheapest test that could validate or kill that assumption before full build
- If that test would take less than 3 days, recommend running it first

## Step 3: Write the brief

### Campaign name and goal
[Name]: [One sentence describing the campaign]
[Goal]: [Specific, measurable. Not "increase awareness".]

### Positioning statement
[For [AUDIENCE], who [PAIN], [BRAND] is the [CATEGORY] that [DIFFERENTIATOR]. Unlike [COMPETITOR/ALTERNATIVE], we [PROOF POINT].]

### Messaging hierarchy
Primary message (one sentence a stranger would understand):
Supporting message 1 (the proof):
Supporting message 2 (the mechanism):
Supporting message 3 (the objection handler):

### Channel plan and sequence
| Channel | Role | Message | Format | Timing |
|---|---|---|---|---|
| [Channel] | [Awareness / Nurture / Conversion] | [Which message] | [Post / Email / Ad] | [Day N] |

### Asset list
[List every piece of content needed. Format: asset type, channel, who produces it, due date.]

### Success metrics
| Channel | Metric | Target | Kill threshold |
|---|---|---|---|
| [Channel] | [e.g. reply rate] | [target] | [below this = pause] |

### What is out of scope
[Explicitly name what this campaign does NOT include. This prevents scope creep.]

### Honest advisor note
[The one assumption the campaign relies on, and what evidence would change the plan.]

## Rules
- No em dashes anywhere
- Never write a goal as "increase awareness" or "build brand" without a measurable proxy metric
- The messaging hierarchy must be in plain English a non-marketer could understand
- The out-of-scope list must include at least 2 items
- Apply brand-voice and audience-profile skills throughout

All 10 skills, install commands, and the recommended build order.

Build order: which skill to install first

# WEEK 1: FOUNDATIONS
# Install these two first. Every other skill in this lead mag builds on them.

# 1. Brand Voice
# Settings > Customize > Skills > Add
# Or: mkdir -p .claude/skills/brand-voice
# Trigger: any content writing for your brand

# 2. Audience Profile
# Settings > Customize > Skills > Add
# Or: mkdir -p .claude/skills/audience-profile
# Trigger: any content or outreach for your brand

---

# WEEK 2: CONTENT PRODUCTION
# Install whichever fits your biggest content bottleneck.

# 3. Content Engine (if you have transcripts to repurpose)
# mkdir -p .claude/skills/content-engine
# Trigger: paste a transcript and ask for posts

# 4. LinkedIn Post Writer (if you post to LinkedIn regularly)
# mkdir -p .claude/skills/linkedin-post-writer
# Trigger: write a LinkedIn post about [TOPIC]

# 5. YouTube Launch Kit (if you publish YouTube videos)
# mkdir -p .claude/skills/youtube-launch-kit
# Trigger: paste a transcript and ask for titles/chapters/description

# 6. Editorial Calendar Builder
# mkdir -p .claude/skills/editorial-calendar
# Trigger: plan my content / build a content calendar

---

# WEEK 3: OPERATIONS
# Install these once the content layer is working.

# 7. Meeting Notes Synthesizer
# mkdir -p .claude/skills/meeting-notes
# Trigger: paste a transcript / clean up these notes / write up the meeting

# 8. Cold Email Personaliser
# mkdir -p .claude/skills/cold-email-personaliser
# Trigger: write a cold email / first line for [PROSPECT]

# 9. Competitive Intelligence Brief
# mkdir -p .claude/skills/competitive-brief
# Trigger: research [COMPETITOR] / brief me on [COMPETITOR]

# 10. Campaign Brief Generator
# mkdir -p .claude/skills/campaign-brief
# Trigger: brief a campaign / plan a campaign for [OFFER]

All 10 skills: one-line reference

Skill	Trigger	What it produces
brand-voice	Any content writing for your brand	Consistent tone, vocabulary, and rules applied to every output
audience-profile	Any content or outreach	Writing calibrated to your specific reader's vocabulary and pain
content-engine	Paste a transcript + ask for posts	3 LinkedIn posts, 2 X threads, 1 SEO article outline
linkedin-post-writer	Write a LinkedIn post about [TOPIC]	Hook type selected, post drafted, content moat checked
youtube-launch-kit	Paste transcript + ask for YouTube assets	5 titles, description, chapters, tags, thumbnail text, pinned comment
editorial-calendar	Plan my content / build a content calendar	4-week calendar with angles, hook types, formats, anti-repetition
meeting-notes	Paste transcript / clean up these notes	Decisions, action items with owners, open questions, summary
cold-email-personaliser	Write a cold email / first line for [PROSPECT]	Waterfall-based opener, subject lines, 3-touch sequence, QA check
competitive-brief	Research [COMPETITOR]	Positioning, G2 complaints, hiring signals, what to say on a call
campaign-brief	Brief a campaign for [OFFER]	Positioning, messaging hierarchy, channel plan, metrics, honest advisor check

Skills that should always load together

Some skills do better work when loaded alongside each other. Claude handles this automatically if your skill descriptions are written correctly, but you can also invoke them explicitly:

# Content engine: always loads brand-voice
"Using content-engine and brand-voice, repurpose this transcript: [PASTE]"

# LinkedIn post writer: loads brand-voice and audience-profile
"Using linkedin-post-writer with brand-voice and audience-profile, write a post about [TOPIC]"

# Campaign brief: loads brand-voice and audience-profile
"Using campaign-brief with brand-voice and audience-profile, brief a campaign for [OFFER]"

# Cold email: loads brand-voice
"Using cold-email-personaliser with brand-voice, write a cold email for [PROSPECT URL]"