The decision-helper skill: stop weighing it and pick one today

You've been going back and forth on the same thing for six weeks. Maybe it's quit or stay, maybe it's raise the price or keep the client. You've read everything, you've made a list, and you're exactly where you started.

The decision-helper skill makes the call. It gets the real options on the table, including the two or three you never listed, then asks which single thing you're optimizing for a year from now. Then it runs a reversibility check, because a decision you could undo in a week doesn't deserve six weeks of thinking. You finish with one recommendation, what you're giving up to take it, the one condition that would prove it wrong, and one move you can make in the next 30 minutes.

It's free and public, one of the 55 skills in our freeskills repo. Nothing to configure.

Claude Code reads skills from a folder on your machine. Put this one there and it turns into a slash command. Three lines:

Restart Claude Code so it sees the new folder. Done. No API keys, no config, nothing to connect.

Type /decision-helper and state the decision in one sentence, then list what you're weighing. It talks back rather than dumping an answer. It'll add options you didn't mention, and it'll drop the ones you've clearly already ruled out but keep listing to feel thorough. Expect it to say that part out loud.

Then it asks the question that does the work: if only one of these could be true a year from now, which one? Money, time, risk, learning, freedom, reputation, keeping the relationship. You have to pick one. After that it goes hunting for the constraint you haven't said, usually a person, a date, or a number you've been talking around.

A good first result: a file in ~/decisions/ with your options, what you're optimizing for, the constraint, the reversibility read, the recommendation, and a decide-by date. It says the first move out loud in the reply so you can do it without opening the file. If that move takes more than half an hour, it isn't the right one.

Give it the honest version, including how long you've been stuck. The length of time is information.

Some choices are cheap to undo and you're spending more time deciding than fixing it would cost. Say so up front and it'll skip the deep analysis.

Most stuck decisions have a person or a date sitting underneath them. This prompt asks for that directly.

Here's the spoken version: what this thing does, and why it changes your week.

1. 1Hook: there's a brand-new free Claude skill that takes you from six weeks of going back and forth to a decision made and a message sent today.
2. 2Mechanism: you tell it the decision you've been circling. It gets the real options on the table, makes you name the single thing you're optimizing for, then surfaces the constraint you hadn't said out loud and checks how reversible each option is. It ends by making the call and naming the first move you can make today.
3. 3Transformation: the loop in your head stops. You know what you picked, what you gave up to pick it, and what would tell you you were wrong. That's a lighter week.
4. 4Close: the skill is free, and there are 55 of them in that repo. Building whatever you just decided on is the part we do inside Claude Code Club, $9 a month. Link below.

Take the decision you've been carrying longest and run it tonight. Do the first move before you close the laptop, even if it's twelve words in a message. That's the part that actually ends it.