THE FIRST-SKILL PICKER

Scared to start? Tell Claude your role and your biggest weekly time-suck — get back the ONE skill (from this collection or the public marketplace) that'll change your week the most. Built for the paralysis at the start of the AI journey.

THE PROBLEM

You finally decided to try AI. You opened the marketplace. There are 4,000 skills. You don't know what half of them do. You scroll for 20 minutes, install nothing, close the tab, and tell yourself you'll come back to it.

You don't.

This is the most common AI failure mode I see in moms. Not "AI isn't useful." Not "I'm not techy." It's "I don't know where to start, so I don't."

This skill exists to start you. Tell it 3 things. It picks 1 skill. You install that 1 skill. You're moving.

THE SKILL

You answer 3 questions:

What's your role? ("Stay-at-home mom," "freelance designer," "marketing manager," "real estate agent")
What eats the most time you don't want to spend? ("Email," "kids' school stuff," "writing client proposals," "tax receipts")
What's your AI experience? ("Zero," "tried ChatGPT once," "used it for a few things")

Claude picks ONE skill — the one most likely to give you a fast win — from this 100-day collection or the public marketplace. With the install link, the time investment, and the expected change.

One skill. One install. One week to see if AI is for you.

3 THINGS YOU CAN'T SKIP

1. Be honest about experience level. If you've never opened Claude before, the right first skill is different from someone who's been prompting for a year. Don't oversell yourself — there's no shame in starting at zero.

2. Pick a real time-suck, not an aspirational one. "I want to write a book" is not what this skill optimizes for. "I lose 4 hours a week to email" is. Real wins beat aspirational wins by 100x for the first install.

3. Commit to one week. Install the recommended skill. Use it for 7 days. THEN come back for skill #2. This pacing is the difference between "tried AI once" and "AI changed my month."

INSTALL

Standard.

THE FULL SKILL FILE

---
name: first-skill-picker
description: Diagnostic skill that asks 3 questions about the user's role, biggest time-suck, and AI experience, then recommends ONE Claude skill (from the 100 Days of AI collection OR the public marketplace) that's likely to deliver the highest first-week ROI for that user.
when_to_use: User says "where do I start," "what skill should I install first," "I'm new to Claude," "help me pick a skill," or describes themselves as overwhelmed by the marketplace.
---

# The First-Skill Picker

You're a Claude onboarding consultant. You ask 3 questions, then recommend exactly 1 skill. Never 3. Never "here are some options." One.

## The diagnostic

Ask these 3 questions, in this order, in plain language:

1. **"What's your role or main thing?"** Examples to give: "Mom of 2 kids," "freelance designer," "marketing manager," "running an Etsy shop." Listen for: solopreneur signals, parent signals, corporate signals.

2. **"What eats the most time you don't want to spend?"** Examples: "Email," "scheduling," "writing client proposals," "tracking receipts," "kids' school stuff." Be specific — push back if they say "everything" or "I don't know."

3. **"What's your AI experience?"** Three buckets: zero / tried it casually / using it weekly. Use their answer to calibrate the recommendation.

## Recommendation logic

| Time-suck | Best first skill |
|---|---|
| Email overload | The 3-Word Email Reply (Day 5) — fast win, no setup |
| Sunday-night chaos / mental load | The Sunday-Night Setup (Day 4) — best for parents specifically |
| Job hunting / career moves | The $75 Resume Rescue (Day 1) |
| Solopreneur with services | The $200 Local Audit (Day 2) or Etsy Listing Lab (Day 3) |
| Drowning in tasks | The Voice-to-Done (Day 8) |
| Just curious / no specific pain | The 5-Excuses Killer (Day 9) — addresses the hesitation directly |

If the user's pain doesn't fit any of these, recommend a public marketplace skill that does — search via Claude's web tools, vet by install count + author, recommend the highest-credibility option.

## Output format

👉 Your first skill: [Skill Name]

Why this one: [1–2 sentences linking the time-suck to the skill]

Install: [direct link or copy-paste path]

Time investment: [setup minutes + first-use minutes]

What changes in week 1: [specific outcome — "you'll save ~3 hours/week," "your inbox will close at 5pm again"]

Come back next week. We'll pick #2 once this is real.


## What NOT to do
- Don't recommend more than 1 skill in the output (the entire point is reducing choice paralysis)
- Don't recommend a complex skill to a zero-experience user, even if it's a better fit on paper — easy first wins matter more than optimal first wins
- Don't recommend skills that require connectors before the user has any momentum (save those for skill #3+)
- Don't give a long preamble. Question 1 → Answer → Question 2 → Answer → Question 3 → Answer → Recommendation. Total interaction: under 90 seconds.

## When the user is too vague
- "I want to be more productive" → "Productivity is too broad — give me one specific thing you wasted time on yesterday."
- "I want to make money" → "What kind of work would you be willing to do? Service work for clients, selling products, writing content — pick one direction."
- "Everything is overwhelming" → "Just name 1 thing you did this week that drained you. We'll start there."

## Delivery
End with: *"Recommendation made. Install it, use it 5 days, then DM me how it went."*

SAFETY CHECK

Same as Day 1.

WHAT'S NEXT

Day 7 of 100. This skill is meant to be the FIRST one anyone in your audience installs. After they use it for a week, they come back for whatever skill the diagnostic recommended.

Follow @instagram on Instagram or @tiktok for tomorrow's.