The launch-plan skill: a dated plan instead of one hopeful post

You built the thing. The plan for selling it is one post on the day it goes live, and then a quiet week where you decide the product was the problem.

A launch is a short window where everything you've got points at one offer, so the sales land together instead of dribbling in. The launch-plan skill builds that window. It picks one launch type for where you actually are, sets the date and the number you're aiming at, lists the minimum you need to make, and writes out what you do each day, from the teasers before it opens to the last-chance message on the final night.

It's part of the free 55 in our freeskills repo. It's also honest with you about size: a launch concentrates demand you already built, and it won't promise a spike your reach can't produce.

A skill is just a folder with a set of instructions inside it. These three lines put this folder where Claude Code looks for it.

Then restart Claude Code, or open a new session. Installing is the whole job. No keys, no configuration.

Type /launch-plan. It asks five things: what you're launching, what stage it's at, how big your audience is right now, whether you have a date in mind, and what the goal is. Give it a real audience number, because that answer changes the plan more than anything else.

Then it recommends one launch type and tells you why that one fits. There are four to choose from: pre-sell, waitlist, a directory launch, or a limited-time deal. It gives you one recommendation, not a menu. If your audience is basically none it says so plainly and swaps the plan for the honest version, usually pre-selling to a small warm list and direct messages, or spending the pre-launch window building a waitlist first.

A good first result: a dated file in ~/launches/ with the launch type and the reason, the open and close dates, one concrete goal number, a short checklist with have or need marked on every line, and a timeline written as dated lines, counting down to launch day and running on through the days after it, with one or two actions on each line. Then it offers to draft any of the messages. Do the first teaser while it's fresh.

Most common starting point. Give it the real audience number, not the one you wish you had.

If it isn't built yet, that's a launch type of its own. A pre-sell tells you whether to spend the next six weekends on it.

The plan is a list of things to write. Ask for the writing in the same session so the assets exist before the date arrives.

Here's what it does to a launch, said out loud.

1. 1Hook: here's exactly what to do on each day of your launch, with one real number to hit. This free Claude skill dates the whole plan for you.
2. 2Step 1: you tell it what you're launching and how many people you can actually reach right now.
3. 3Step 2: it picks one launch type that fits that size, and tells you why.
4. 4That number decides everything that comes after it.
5. 5Step 3: it sets the date and one concrete goal number.
6. 6It also knows which day most of the money comes in.
7. 7Step 4: it works backwards from that date into a day-by-day plan: what you post before it opens, what goes out on the day, and what you send on the last night when most of the sales land.
8. 8Transformation: you stop guessing what to do each day of your launch. You open the file, do the two things listed under today's date, and close the window on the date you said you would.
9. 9Close: it's free, and it's one of 55 sitting in that repo. Building the audience that makes a launch worth running is Claude Code Club, $9 a month. Link below.

Tonight: run it, pick a date you're willing to hold, and write the first teaser. Deadlines you extend teach people your deadlines aren't real, and the next launch pays for it.