DEV Community: Joe Kurt

How to Write SKILL.md Files in Claude for Non-Coding Work (With a Real Test)

Joe Kurt — Mon, 09 Nov 2026 00:38:29 +0000

A practical breakdown of the 4 blocks inside a SKILL.md file that get clean client and business work in one turn.

If you have hit Claude's 5-hour session limit halfway through an afternoon of ordinary work, there is a good chance the problem was not the size of your project. It was the number of messages it took to get a usable draft.

Watch how most of us use Claude when we are in a hurry. We paste rough notes from a client call and ask for a proposal. Claude comes back with 900 words of polite corporate filler and a single flat price. So we send a second message asking for three pricing tiers. Then a third message asking it to sound less stiff. Then a fifth message reminding it to list what is out of scope so we do not get stuck doing free revisions.

By message ten, every single correction has been re-read into the context window on every turn. You spent twenty minutes babysitting a chat box and burned five times more session tokens than the job actually needed.

Developers solved this months ago for code by moving repeatable instructions into small SKILL.md files. Yet almost every public SKILL.md catalog is still packed with React, SQL, and Docker tools, while the everyday work that eats up most people's week (scoping client proposals, turning messy call notes into action lists, stress-testing a launch plan) gets left in scratchpad prompts.

Over the past few weeks I moved my own non-coding business tasks into portable SKILL.md files that work across Claude Code, Claude Cowork, and plain Claude Projects. Here is how the file format actually works under the hood, the four blocks every reliable skill needs, and an unedited test run from a 24-word prompt.

If you would rather have the files than build them, I keep mine here: https://livefree1.gumroad.com/l/chpvtb

How Claude actually reads a SKILL.md file

A skill is just a folder with a lowercase hyphenated name containing a plain Markdown file named SKILL.md:

skills/
  client-proposal-scope-builder/
    SKILL.md
  meeting-transcript-to-action/
    SKILL.md
  premortem-blindspot-finder/
    SKILL.md

What makes SKILL.md smarter than pasting a giant mega-prompt into every chat is how Claude loads it.

At the start of a session in Claude Code or Claude Cowork, Claude does not read the entire body of every skill you have installed. It only scans the short YAML header (name and description) at the top of each file. When your request matches one of those descriptions, Claude pulls that single SKILL.md file into context, follows its rules, and leaves the rest of your skills untouched.

If you use standard Claude on the web or phone instead of the terminal, the exact same SKILL.md file drops straight into the Project Knowledge panel inside a Claude Project (or can be pasted into any chat window when you need it).

The 4 blocks that make a SKILL.md file work on the first try

Early on, half of the skill files I wrote still gave me mushy, generic answers. Fixing them came down to stripping out vague wishes ("be clear and persuasive") and building every SKILL.md out of four specific parts.

1. The YAML trigger header

The top of every SKILL.md starts and ends with three dashes (---) and holds two required fields: name and description.

Because Claude decides whether to load your skill based purely on those two lines, write the description around two things: what the output looks like, and the exact moment in your workday when Claude should use it.

Bad description:

Helps write better proposals for clients.

Working description:

Turns rough discovery notes into a clear 1-to-2 page proposal with 3 options and scope protection. Use right after a client call or inquiry to send a clean, fast proposal while momentum is high.

2. A short intake guard

When you hand an AI model incomplete notes, its default instinct is to guess the missing numbers so it looks helpful. In real client or strategy work, made-up timelines and imaginary prices are worse than useless.

Right under the header, add a 3-line intake check telling Claude which two or three facts must be present before it starts writing. If you forgot to mention your base price or target timeline, Claude stops and asks you one short question instead of inventing a number.

3. A numbered execution sequence

Long paragraphs of instructions invite the model to cherry-pick whichever rules feel easiest. A short numbered sequence forces a consistent build order every single time.

For example, in my proposal skill, the sequence never changes:

Open with the client's situation and goal in plain words so they see you actually listened on the call.
Present three priced options (1x Essential, 1.5x Recommended Complete, and 2.2x Fast-Track / Done-for-You) so the client compares how to work with you rather than whether to hire you.
Add an explicit Included vs. Out-of-Scope section (revision cap, hosting, third-party tools) to stop unpaid scope creep before it starts.
Close with milestone dates, deposit terms, and a one-line reply to accept, followed by a calm cover email.

4. A falsifiable quality gate

Telling Claude to "sound human" fails because you cannot point to a sentence and prove objectively whether the rule was followed. Good rules are falsifiable: you can check pass or fail in five seconds by eye.

At the bottom of every SKILL.md, I include three hard checks:

Ban specific dead-giveaway filler words (delve, tapestry, landscape, game-changer, seamless, leverage, crucial, foster, moreover).
Ban the classic AI contrast tic ("It is not just X, it is Y").
Require concrete numbers, dates, or named deliverables in every bullet instead of general advice.

A complete copy-pasteable SKILL.md file

Here is the full plain-text SKILL.md file for the client proposal workflow so you can inspect or copy it directly:

---
name: client-proposal-scope-builder
description: "Turns rough discovery notes into a clear 1-to-2 page proposal with 3 options and scope protection. Use right after a client call or inquiry to send a clean, fast proposal while momentum is high."
---

CLIENT PROPOSAL & SCOPE BUILDER

WHEN TO USE THIS SKILL
Paste your messy call notes, client email, or project idea to produce a ready-to-send proposal in 5 minutes.

STEP 1: FAST INTAKE CHECK
Ask for (if not already in notes): (1) what the client actually wants achieved, (2) rough timeline, and (3) your target price for the standard option.

STEP 2: EXECUTION PROTOCOL
Write in plain, direct human English (grade 7-9 clarity, concrete numbers, zero corporate cliches). Execute in order:
1. Open with SITUATION & GOAL in the client's own words to prove you listened before talking about deliverables.
2. Present 3 CLEAN OPTIONS (Option 1: Essential / Core fix at 1x price, Option 2: Recommended Complete Outcome at ~1.5x price, Option 3: Fast-Track / Full Done-For-You at ~2.2x price) so the client compares how to work with you rather than whether to hire you.
3. Write explicit SCOPE BOUNDARIES ('What Is Included vs. Out of Scope') including revision cap, turnaround window, and what counts as an add-on to prevent unpaid scope creep.
4. Close with TIMELINE, PAYMENT TERMS, AND 1-STEP ACCEPTANCE so they can say yes today.

STEP 3: REQUIRED OUTPUT FORMAT
Deliver the result cleanly using this exact structure:
1) READY-TO-SEND PROPOSAL DOCUMENT and 2) SHORT COVER EMAIL to send alongside the proposal.

QUALITY GATE (VERIFY BEFORE RETURNING)
- Zero dead-giveaway AI words (delve, tapestry, landscape, game-changer, seamless, leverage, crucial, foster, moreover).
- Zero "Not just X, it's Y" phrasing or empty cheerleading.
- Every recommendation uses concrete actions, real numbers, or clear dates rather than vague advice.

Testing it on a real 24-word note

To test how little input a structured SKILL.md file actually needs, I ran that skill on a single 24-word note:

Client: Karim owns a boutique cafe, wants a 3-page website with menu + WhatsApp table booking ready in 2 weeks. My standard price is $800.

In one turn (about fifteen seconds, zero follow-up edits), here is what came back:

Three distinct packages priced cleanly from the $800 anchor: Essential Website ($800) with 3 pages, mobile layout, WhatsApp click-to-chat button, and 2 revision rounds; Complete Cafe Website ($1,200, Recommended) adding categorized food photography, Google Maps, custom pre-filled WhatsApp message, and 3 revision rounds; and Premium Done-for-You ($1,760) adding priority scheduling, image optimization, Google Business Profile link setup, and a recorded handover walkthrough.
An automatic Out of Scope block protecting the builder from the exact traps cafe clients usually assume are free: live table-availability databases, online payment processing, food photography, domain and hosting renewals, and post-launch maintenance.
A day-by-day 14-day schedule (Days 1-2 kickoff, Days 3-6 structure, Days 7-10 build, Days 11-12 revisions, Days 13-14 handover), a 50% / 50% payment split, and a calm 6-sentence cover email with zero "I hope this email finds you well" filler.

How to install or adapt 3 open-source examples

If you want to test this setup without writing your first SKILL.md from a blank screen, I put three of my everyday non-coding business skills into a free MIT-licensed GitHub starter repository (client-proposal-scope-builder, meeting-transcript-to-action, and premortem-blindspot-finder):

https://github.com/joekurt120/claude-work-skills

If you use Claude Code, Cursor, Codex, or Windsurf, you can install all three in one command with the open skills CLI:

npx skills add joekurt120/claude-work-skills

And if you never touch a terminal, just open the skills/ folder on GitHub, grab whichever SKILL.md file matches a job you repeat every week, and drop it into a Claude Project. Swap my steps for the rules you normally find yourself typing on message four or five of a messy chat, and see how much quieter your sessions get.

Want the files instead of building them?

I keep the skill files I actually use, plus the context file and the writing rules file, in one pack here:

https://livefree1.gumroad.com/l/chpvtb

The launch checklist is free if you just want somewhere to start:

https://livefree1.gumroad.com/l/sekiij

Claude Skills for Client Work: The 10 I Use Every Week

Joe Kurt — Fri, 06 Nov 2026 23:47:08 +0000

Most skill folders get opened once and never again. Here is what survived a year of real proposals, calls, and invoices, plus the two files that make all ten of them work.

I have deleted more Claude skills than I have kept.

That is not a brag, it is the most useful thing I learned about them. A Claude skill is just a folder with a file inside called SKILL.md. Claude reads the folder name and a one line description at the start of every chat, then opens the file when your request matches.

The problem is not the format. It is that most of us build skills for work we do twice a year, then wonder why the folder sits there untouched.

If you do client work, freelancing, consulting, or run a small agency, this is the filter that fixed it for me: if I have not done the task at least twice this week, it does not become a skill. It stays a prompt.

Below are the ten that survived that filter, the two files they all depend on, and one complete SKILL.md you can copy and change.

If you would rather have the files than build them, I keep mine here: https://livefree1.gumroad.com/l/chpvtb

Why most Claude skills get used once

Three reasons, and I have made all three mistakes.

The description is too wide. Write "use this for business writing" and it fires on every message, which means it helps with none of them. A good description sounds like something you would say out loud to a colleague.

There is no context. A skill that does not know who you are, what you sell, or who you sell to will produce confident generic work. That is the actual reason output sounds like it came from a machine.

It was built for an imaginary task. You built it because it sounded useful, not because you were tired of repeating yourself.

Fix those three and the format starts working.

The two files that make the other ten work

Build these before any of the skills below. Everything else inherits from them.

Your context file

Six short sections. Who you are and what you do. Who you serve and what hurts them. What you sell and what it costs. What proof you actually have. Your current 90 day goal. How you want Claude to talk to you.

Fill it in once. It takes about four minutes, and you can have Claude interview you to do it. Paste the whole file into Claude and say "interview me five questions at a time to fill this out cleanly."

Your writing rules file

Mine is a list of words and sentence patterns I refuse to publish. It started with eight words. It has about sixty now, because every time I caught Claude writing something I would never say, I added it.

Put this file in your project instructions and every skill below obeys it automatically. You stop fixing tone in nine different places.

Here is a version you can copy and adapt.

Write the way an experienced professional talks to another person.
Use plain words and short sentences. Mix a six word sentence with a longer one when it helps the rhythm.
Say the point in the first sentence. Do not warm up for three lines before saying something real.
Use real numbers instead of vague claims. "Cuts a 45 minute draft to 6 minutes" beats "saves a lot of time."
Never use these words: delve, embark, unleash, unlock, elevate, skyrocket, supercharge, harness, foster, spearhead, showcase, navigate, streamline, tapestry, landscape, testament, cutting-edge, seamless, robust, vibrant, pivotal, crucial, realm, beacon, synergy, paradigm, holistic, myriad, plethora.
Never use these openers: "In today's fast-paced world", "Let's dive in", "Here's the kicker", "Moreover", "Furthermore", "At the end of the day", "Whether you're a X or a Y".
Never write "Not just X, it's Y."
Never open with "Great question!" or close with "You've got this!"
Before you finish, read it back. If a real person would not say it across a coffee table, rewrite it.

The ten Claude skills for client work

Here they are at a glance.

Skill	What you paste in	What you get back
Client brief builder	Messy client messages	What they want, what is missing, what to charge
Proposal with three prices	Call notes, target price	Three options plus a short cover email
Scope creep check	A request from a client	Inside or outside scope, plus the reply
Meeting recap	Transcript or rough notes	Three bullets, action table, recap email
Pricing check	Your price, monthly target	The arithmetic after fees, and a verdict
Difficult email	The situation	Two tones, plus what to leave out
Sales page copy	Offer, price, buyer, proof	Every section from headline to FAQ
Pre-mortem	Your plan	Ranked failure list and a 48 hour test
Notes into a procedure	You explaining a task	A checklist someone else could follow
Weekly review	Five rough bullets	What shipped, the bottleneck, next week's three

Now the detail on each.

Client brief builder

A client sends three paragraphs of half formed thoughts. This returns what they actually want, what is missing, what it should cost, and what you need before you start.

The part I underestimated: it writes the questions you should ask back. Half of scope creep is a brief nobody pushed back on.

Proposal with three prices

One option and people negotiate. Three and they choose. Give it your call notes and a target middle price.

Ask for a Basic, a Recommended, and a Premium. Put the one you want in the middle. Price it so the middle one looks like the obvious decision, because that is what it is.

Scope creep check

Someone asks for "one small addition." This checks the request against what was agreed and tells you plainly whether it is inside or outside scope. Then it drafts the reply in a tone that keeps the relationship.

It also tells you what not to add to that email. That line has saved me more money than the email has.

Meeting recap

Three bullets at the top, because that is all anyone reads. Then a table of actions, owners and deadlines. Then the things nobody said out loud but everybody noticed. Then the recap email.

That third section is the one clients reply to. It shows you were listening.

Pricing check

You name a price and a monthly target. It does the arithmetic after fees and shows whether your pricing can reach that number at all.

I have killed two offers with this. Both would have taken months and could never have worked at the price I had in my head.

Difficult email

Asking for money. Saying no. Holding a boundary on a client who keeps pushing.

Ask for two versions, one warm and firm and one direct, plus a note on what to leave out. What you leave out is usually what starts the argument.

Sales page copy

Section by section, from the headline down to the FAQ. Give it your offer, your price, your buyer, and the real proof you have.

It will not invent proof. That is deliberate. If the proof section comes back thin, that is information about your offer, not about the skill.

Pre-mortem

Before you commit weeks to something, ask it to be ruthless. You get a ranked list of how this fails, what you have assumed without evidence, and one cheap test you can run in 48 hours.

Run it on launches, hires, and pricing changes. It costs ten minutes and it is the cheapest insurance you will ever buy.

Notes into a procedure

Talk through how you do something. It comes back as a checklist someone else could follow without asking you questions.

This is how you stop being the bottleneck. It is also how you find out your own process has a hole in it, which is uncomfortable and useful.

Weekly review

Five rough bullets on what you shipped, your numbers, and what felt slow. It separates real output from motion, names your single bottleneck, and sets three priorities for next week.

Fifteen minutes on a Friday. The bottleneck it names is almost never the one you would have guessed.

What a real SKILL.md file looks like

Here is the proposal skill, complete. Copy it, change the parts in brackets, save it as SKILL.md in a folder named client-proposal-builder.

---
name: client-proposal-builder
description: Turns rough call notes into a client proposal with three priced options and a short cover email. Use when someone asks for a proposal, a quote, a scope, or pricing for client work.
---

CLIENT PROPOSAL BUILDER

BEFORE YOU START
Read 00_MY_CONTEXT_MEMORY.md if it exists in this project. It has who I am, what I sell, and who I sell to. Never ask for anything already written there.

ASK FOR THESE IF THEY ARE MISSING
a. What the client actually wants, in their own words
b. The budget range they mentioned, or your target middle price
c. Any deadline
d. What "done" looks like

DO THIS IN ORDER
1. Summarise their problem in two sentences.
2. List what is included and what is explicitly not included.
3. Build three options.
   - Basic: the smallest version that still solves the problem
   - Recommended: what you would actually sell them
   - Premium: the same work plus the thing they will wish they had asked for
4. Price each one. Put the recommended option in the middle.
5. Write the cover email. Four sentences. No greeting padding.

OUTPUT FORMAT
The proposal in plain text with clear section labels, then a divider, then the cover email.

RULES
Obey the banned words in 01_HUMAN_VOICE_RULES.md.
Never invent proof, testimonials, or results I have not given you.
If the notes are too thin to price honestly, say so and ask the two questions that would unblock you.

That is the whole format. Name, description, instructions, output shape, rules. The description is what makes it fire, so spend your time there.

Why your skill never fires, and how to fix it

The description is a summary, not a trigger. "Helps with proposals" is a summary. "Use when someone asks for a proposal, a quote, a scope, or pricing" is a trigger.

It fires on everything. Write it as if you are describing the exact moment, not the general topic.

The file is too long. If your SKILL.md is longer than about 300 lines, move the detail into a second file and tell Claude when to open it. Long files get half followed.

You never tested it against things it should ignore. Give it five requests it should stay silent on. If it wakes up for those, the description is too wide.

How to set this up in one afternoon

Make the context file first. Make the writing rules second. Then build one skill and use it on real work the same day.

If you have not used it again by Friday, delete it.

A folder of ten skills you use beats a library of a hundred you do not. I learned that by building the hundred first.

Frequently asked questions

Do Claude skills work on the free plan?

Skills live in Claude Projects, Cowork, and Claude Code. You need a paid plan for Projects. The files themselves are plain text, so you can paste them into any chat and they still work, you just lose the automatic trigger.

What is the difference between a skill and a prompt?

A prompt you retype. A skill sits in a folder and Claude opens it when your request matches, without you asking. That automatic part is the whole point.

How many skills should I have?

Fewer than you think. Ten good ones beat fifty you forget. If you have not used one in a month, it is clutter.

Can I use the same skills in ChatGPT or Gemini?

The file is plain text, so yes, you can paste it in. What does not carry over is the automatic loading, because that is specific to how Claude handles the folder format.

Should I download skills from other people?

Read the SKILL.md before you install anything. It is a plain text file, so check whether it tells Claude to call a URL or write files outside its own folder. Most are fine. Not all.

Want the files instead of building them?

I keep the skill files I actually use, plus the context file and the writing rules file, in one pack here:

https://livefree1.gumroad.com/l/chpvtb

The launch checklist is free if you just want somewhere to start:

https://livefree1.gumroad.com/l/sekiij

How I Use Reusable Claude Skill Files to Run My Daily Work

Joe Kurt — Fri, 30 Oct 2026 04:05:36 +0000

Most people who use Claude every day have a messy folder or spreadsheet full of saved prompts somewhere on their laptop.

I had one too. It had over a hundred prompts I copied from articles and threads, and I almost never opened it. Every morning I still opened a blank chat, typed a rough paragraph explaining what I needed, and spent the next twenty minutes arguing with the output.

One draft came back sounding like a stiff corporate press release. Another forgot my pricing model. By the twelfth message in the same thread, two annoying things always happened at once: Claude started losing track of instructions from the top of the chat, and I burned straight through a huge chunk of my 5-hour session limit.

Most people don't realize how token math works inside a long AI chat. Every time you send message fifteen in a messy thread, Claude re-reads messages one through fourteen all over again. Fixing a weak prompt through twenty rounds of trial and error is the fastest way to hit the usage wall by mid-afternoon.

Everything got easier the week I threw out my prompt list and switched to small, single-purpose Markdown skill files.

Why Prompts Break Down (and Plain Files Work)
A normal prompt is just a wish. You type "write a client proposal from these notes" and force Claude to guess the structure, the pricing layout, the revision boundaries, and the tone all at the same time. When it guesses wrong, you spend six more messages fixing it.

A skill file works like a repeatable checklist for one specific job you do every week.

Instead of keeping everything inside one giant instruction block, I split my setup into two layers:

First, two tiny root files sit permanently inside my Claude Project (or local Claude Cowork folder). One holds my work context: what I do, who buys from me, my offers, real proof numbers, and my 90-day goal. The second holds my plain-English writing rules and blocks sixty common AI clichés so nothing comes back sounding robotic.

Second, I keep twenty small Markdown files organized by the actual parts of a workweek: writing, client work, offers and sales, strategy decisions, and daily operations. When I finish a call or need to price a project, I drop in that one small file, paste my rough notes, and get a finished draft on turn one.

Three Skill Files You Can Copy Today
You do not need any coding setup or special plugins to run this. You can save each block below as a plain .md text file on your computer, drag it into Claude Projects or Claude Cowork, or simply paste it above your rough notes in a fresh chat.

Here are the three workflows that save me the most time every single week.

Turning Messy Call Notes Into a Real Action Plan
Right after a client call, a discovery chat, or a team meeting, you usually have forty lines of messy bullet points or a raw transcript. If you just ask Claude to summarize it, you get a polite paragraph that hides who actually owes what by Friday.

Save this as 08_call_to_action.md and paste your raw notes underneath it:

Purpose: Turn raw call notes or a meeting transcript into clear decisions, owners, hidden signals, and a short recap email.

Intake check: Run right away on the pasted notes. Only pause to ask who I am in the transcript if speaker names are unclear.

Execution steps:

Separate real decisions that both sides agreed on from ideas that were only casually mentioned.
Build a four-column action list showing the task, the owner, the exact deadline, and what is waiting on it.
Flag between-the-lines signals: budget clues, unspoken hesitations, or open questions nobody answered on the call.
Draft a plain-English recap email under 170 words locking in what was agreed and who does what next.
Quality check: Use short paragraphs, zero corporate buzzwords, and concrete dates instead of vague words like "soon."

Turning Discovery Notes Into a 3-Option Proposal
Sending a single flat price after a good client call forces the buyer into a yes-or-no corner. Even worse, sending a loose one-page quote without clear boundaries almost always leads to unpaid extra revisions three weeks later.

Save this as 06_proposal_scope.md and feed it your call bullets plus your target middle price:

Purpose: Turn rough project notes into a clean proposal with three clear options and firm scope boundaries.

Intake check: Ask me for any missing piece from these three items before writing: what outcome the client wants, the rough timeline, and my target price for the standard middle option.

Execution steps:

Open with the situation and goal written in the client's own plain words so they see immediately that I listened.
Build three clear tiers: Option 1 covers the essential core fix at 1x price, Option 2 covers the complete recommended outcome at roughly 1.5x price, and Option 3 covers fast-track or full done-for-you delivery at roughly 2.2x price.
Write an explicit Scope Boundaries section spelling out what is included, the exact revision cap, turnaround times, and what counts as a paid add-on.
Close with the timeline, payment split, and a simple one-step reply so they can approve today, plus a four-sentence cover email.
Running a Ruthless Pre-Mortem Before You Spend Money or Weeks Building
Chat assistants are trained to be agreeable. If you pitch a weak project idea, a risky hire, or an underpriced service, default Claude will cheerfully tell you it sounds promising and help you waste six weeks building it.

Before I commit real hours or cash to a new plan, I run 13_premortem.md so Claude pushes back like a skeptical partner:

Purpose: Assume my plan failed six months from today and expose the real reasons before I spend time or money.

Intake check: Ask me to state my plan, how many hours or dollars I am betting on it, and what success looks like in ninety days.

Execution steps:

Fast-forward six months and assume this plan flopped completely. Do not soften the feedback or flatter me.
Rank the five most likely causes of failure based on how often they happen and how much damage they cause (such as having no distribution plan, pricing below platform fees, or spending weeks building before testing demand).
Point out wishful assumptions in my plan that I am treating as guaranteed without proof.
Give me one cheap 48-hour test I can run this week to prove strangers actually want this before I build the whole thing.
How to Keep Your Setup Fast and Clean
If you build your own folder like this, keep one simple rule: never dump all twenty workflows into a single chat at the same time.

Keep only your short context memory and your voice rules loaded all the time. Pull in individual workflow files only for the exact task sitting in front of you, and start a fresh chat whenever you switch from writing a client proposal to reviewing weekly numbers. Your outputs stay sharp, your voice stays consistent, and your usage meter barely moves.

If you want to build this yourself, start with the three files above and add one new .md file every time you catch yourself typing the same instructions twice.

And if you want my complete folder already built and ready to drop into Claude Projects, Claude Cowork, or a 1-click offline browser dashboard, I packaged all 20 workflow files plus the 2 core memory and human-voice files into The Claude Work and Business OS (or grab my free 48-Hour Launch Checklist first if you are shipping a new offer this week).

How to Launch Any Offer to a Small Email List (My 5-Email Sequence)

Joe Kurt — Tue, 27 Oct 2026 23:55:28 +0000

Almost everybody makes the same mistake the first time they launch a new offer online.

You spend weeks putting together a service package, a digital product, a mini-course, or a small tool you know can help people. Launch morning arrives, you type up one quick announcement email in twenty minutes, hit send, and check your inbox or Stripe dashboard every ten minutes.

Nothing happens. Maybe one polite friend replies to say congrats. By evening, you assume nobody wants what you built.

Most of the time, nothing is wrong with your offer at all. One single announcement email rarely sells anything, whether you are pitching a $29 template or a $1,500 consulting sprint. Half of your subscribers opened your email on their phone while waiting in line for coffee, meant to look later, and completely forgot. The rest wanted to see a concrete example of how it works before spending money.

At the same time, nobody wants to become that marketer slapping fake midnight countdown timers onto an offer that isn't actually closing.

Here is the 5-email launch sequence I use to launch any product, service, or online offer to a small email list over five days without fake urgency.

Why five launch emails beat one announcement

Even on a small list of 40 or 200 subscribers, people make buying decisions at different speeds. A few readers buy on Day 1 the moment they recognize the problem. Most people wait until Day 2 or Day 3 when they can see how your process works behind the scenes and get their doubts answered. And a huge group of buyers wait until the final email simply because work got busy mid-week.

Spreading your launch across five short, helpful emails lets every reader buy at their own pace without feeling spammed.

Day 1: Start with a real problem and a free quick win

Skip the loud "I'm thrilled to announce!" intro. Nobody wakes up hoping to read a company press release.

Open your first email by describing the exact everyday bottleneck your client or customer hates dealing with, then teach one useful piece of the fix right inside the email for free. Once the reader gets something practical out of opening your message, introduce your full offer or service in the last two short paragraphs for anyone who wants the complete solution.

Day 2: Show how it works behind the scenes

Vague promises make smart buyers keep their wallet closed. Instead of telling people your offer will "save hours every week," show them one real piece of it in action.

If you sell a template or course, walk through one real page or lesson. If you sell a service or consulting package, walk through step one of how you audit a client's setup. When someone can picture how your work fits into their week, saying yes gets much easier.

Day 3: Answer the skeptical question in their head

Every reader who clicked your link on Day 1 and left without buying had one quiet question holding them back. Usually it is "Will this work for my specific situation?", "How much time will this take me?", or "Why pay for this when free advice exists?"

Write one short email answering that exact question straight up. Right underneath it, add three honest bullets on who should NOT buy it. Telling someone to keep their money if it is not a fit builds more trust in ten seconds than a full page of sales hype.

Day 4: Compare the slow manual way to the faster way

Skip big adjectives and show simple math instead. Walk through what the old way costs in time or headaches (like spending four hours every Monday doing client reporting or rewriting emails from scratch) versus what happens once your system, tool, or service is in place.

Day 5: Send a calm last call (copy-paste template)

Roughly a third of small-list sales and client bookings come in on the fifth email, not because of fake scarcity, just because busy subscribers finally clear their open tabs. Tell people plainly that this is your last note about the offer before returning to regular content tomorrow.

Here is the exact Day 5 launch email template you can copy and adapt for any product or service:

Subject: Last note on Offer Name

Quick heads-up before I get back to regular posts tomorrow.

Earlier this week I shared [Offer Name], built to help you [plain-English outcome, e.g., onboard new clients or write launch emails in 20 minutes] without [main headache, e.g., starting from a blank page or losing weekends to admin work].

Here is what is included:

[Part 1 of your offer or service and what it solves]

[Part 2 and how fast they can use it]

[Bonus, walkthrough, or quick-start step]

If you already grabbed a spot or picked up a copy this week, thank you. If you meant to check it out earlier and got busy, here is the link one last time:
[Your Link]

Regular posts pick back up tomorrow.

Try this 5-day rhythm on the next product, service, or project you put out. You keep every bit of trust with your audience, and you stop leaving two-thirds of your sales on the table after Day 1.

P.S. Want to launch your next offer without writing everything from scratch? You can grab my **free 48-Hour Launch Checklist* (and check out The Launch Email Vault with all 30 plain-English launch, follow-up, and client-ready email templates, plus The Claude Work & Business OS) right here: livefree1.gumroad.com*

How to Stop Claude From Sounding Like AI (My Claude Project Instructions)

Joe Kurt — Mon, 26 Oct 2026 03:14:53 +0000

Every time I opened a new chat in Claude, I used to waste the first ten minutes fixing the exact same problem.

I would explain what I do, tell it who I was writing for, and ask for a normal, conversational tone. Five seconds later, it still handed me stiff paragraphs packed with words nobody says in real life, like "delve," "tapestry," "game-changer," and "in today's fast-paced world." Then I spent twenty minutes rewriting the whole draft by hand.

If you want to make Claude write like a human on the first pass, better one-line prompts are not the answer. Putting two short files inside your Claude Project instructions (or your Claude Cowork folder) fixes it permanently.

Why Claude sounds stiff out of the box

By default, Claude knows nothing about your work, your customers, or how you talk. When it lacks specifics, it fills the gap with safe corporate filler. Giving it persistent context and a hard list of banned patterns stops that habit before it starts.

1. My Claude Project memory file

Instead of re-introducing yourself in every new chat, create one plain text file called 00_MY_CONTEXT_MEMORY.md and upload it to your Project knowledge. Keep it to six short paragraphs:

What you do and your real background in one plain sentence.
Who buys from you (or reads your work) and what problem has them stuck right now.
Your offers or services, their prices, and the concrete result they deliver.
Real numbers, past wins, or experience you can prove (plus a strict rule never to invent fake stats).
Your main business goal for the next 90 days.
Your working ground rules: "Write in plain English. Skip flattery. Tell me straight when my idea has a hole in it."

If you hate staring at a blank page, paste those six points into Claude and ask it to interview you three questions at a time, then write the finished file for you. It takes four minutes once, and you never have to repeat your background again.

2. My Claude banned words list (copy and paste)

Telling an AI model to "sound natural" rarely works because "natural" means different things to different people. Giving your Claude custom instructions a strict blocklist works right away.

Drop this block into a second file called 01_HUMAN_VOICE_RULES.md inside your Project:

Plain-English Voice Rules:
Write at a clear Grade 8 reading level. Keep paragraphs between 1 and 3 sentences. State the main point in the very first sentence with zero throat-clearing. Do not use long em-dashes.

Never use these words: delve, unleash, unlock, elevate, supercharge, harness, foster, leverage, streamline, tapestry, landscape, testament, game-changer, cutting-edge, seamless, robust, pivotal, realm, synergy, holistic.

Never use the "not just X, it's Y" sentence pattern (for example: "This isn't just a checklist, it's a system"). State plainly what something is and what it does.

Turning repeat work into reusable Claude skills

Once those two files sit inside your Project, drafts immediately sound cleaner and closer to your real voice. From there, anytime you catch yourself repeating the same job three or four times a week (writing client proposals, cold emails, landing pages, or weekly plans), save the exact steps as a reusable SKILL.md workflow file right beside them.

Set up just the memory sheet and the banned-word block above today, then test them on your very next email or post. You will notice the difference on the first reply.

P.S. If you want my complete ready-to-use setup without building it from scratch, I put both full Core Memory and Human Voice files, 20 installable Claude SKILL.md workflows for everyday writing, client work, sales, and operations, plus a 1-click offline prompt dashboard (along with my free 48-Hour Launch Checklist) right here: livefree1.gumroad.com