The skill library

Prompt skills:
the moves, not the magic.

There are no magic prompts — there are repeatable moves. This page is our working library: the Yap Method kit, plus the best techniques we've learned from two of the sharpest AI teachers publishing today, Peter Yang (Creator Economy) and Sabrina Ramonov (sabrina.dev) — every prompt rewritten by us for a brain with 43 tabs open. Tap any prompt to copy it.

Start here

The Yap Method core kit

Five prompts that run the whole system. If you only ever save one page of this site, save this one.

Read this before you copy anything — the postcard rule rides along with every prompt on this page: fake names — Person A, Company B, Project X. Nothing you wouldn't write on a postcard. No medical, legal, or money details.

🎙️ The master prompt

Sets up any chatbot as your ADHD-friendly helper — one question at a time, no walls of text.

You are my ADHD-friendly helper. Rules: keep every answer short — no walls of text, one question at a time. I'm going to YAP now. Just listen — don't solve, don't judge — until I say DONE. Then MAP my yap into five short piles: Tasks, Worries, Ideas, Feelings, Questions. Max 5 bullets per pile. No big plan. Then help me TAP: give me exactly ONE tiny physical action that takes under 10 minutes. If it feels too big I'll say SMALLER and you shrink it. When I stop working, make my SNAP card: what I did, where I stopped, my next tiny tap, and one thing future-me needs to remember. Your job is to make my next step smaller — never bigger.

🚨 Emergency overwhelm

For the moment the room is spinning. No questions, no plan — one thing.

I'm overwhelmed right now. Don't ask me questions. Tell me one tiny thing to do in the next 5 minutes. Two sentences max.

🤝 Body double

Borrowed structure on demand. (People with ADHD swear by body doubling; it's not clinically proven.)

Be my body double for 10 minutes. I'm working on Project X. I'll check in every few minutes — when I do, cheer briefly and tell me to keep going. Don't give new ideas. At the end, ask what I got done.

💾 Restart from a save card

The comeback prompt. Feeds your last SNAP card back in and warms you up.

Here's my last SNAP card: [paste]. Warm me up — remind me what I was doing, where I stopped, and my next tiny tap. Short answer.

🐇 The anti-rabbit-hole brake

For when the AI chat itself becomes the rabbit hole.

I've been talking to you a while without doing anything. Stop giving me new ideas. Pick one action from what we already have, give it to me, then don't reply until I say DONE.

Skills we learned from

Peter Yang — structure that survives your brain

Peter Yang writes the Creator Economy newsletter (creatoreconomy.so) and teaches prompting like a craft. These six moves are his frameworks, rebuilt in our words for ADHD entrepreneurs. Go subscribe — the originals are worth it.

🎭 Role + goal, always

His non-negotiable opener: tell the AI who it is and what winning looks like, before anything else. Vague in, vague out.

You're a patient operations manager for a one-person business run by someone with ADHD. Your goal: get me from "everything is on fire" to one finished task today. Follow these steps carefully: 1) Ask me what's on fire. 2) Pick the one with the nearest real deadline. 3) Give me its first 10-minute step. Ask me for more information if you need it.

🪞 Feed it your own best work

His biggest quality lever: show the AI your proven output, not a description of it. New drafts come back sounding like you on a good day.

Here are 3 things I made that actually worked: [paste your best email / post / proposal]. Study what they have in common — tone, length, rhythm. From now on, everything you draft for me matches that. Confirm what you noticed in 3 bullets, then wait for my next task.

🧰 The prompt that writes prompts

Meta-prompting: when you don't know how to ask, make the AI build the ask. Perfect for the "I know what I want but not how to say it" fog.

I need a reusable prompt for this recurring task: [describe the task messily — messy is fine]. Write me that prompt. Give it a role, a goal, numbered steps, one example, and a line telling the AI to ask questions when unsure. Keep the whole prompt under 150 words so I'll actually reuse it.

⚖️ The decision copilot

For decisions you've been circling for weeks: facts first, hard questions second, a one-page record third — so the decision stops living in your head.

Help me make one decision: [decision]. Step 1: ask me, one at a time, for the facts you need. Step 2: challenge my assumptions — ask why up to five times. Step 3: write a half-page decision note: the situation, your recommendation, the strongest alternative, and my next tiny step. I have ADHD — keep each turn short.

✅ The pass/fail checker

His QA trick: AI can't reliably grade 1-to-10, but it can check pass/fail. Turn "is this good?" into a checklist.

Here's my draft: [paste]. Check it pass/fail against this list, one line each: 1) Is the ask clear in the first two sentences? 2) Under 200 words? 3) Zero jargon a stranger wouldn't know? 4) Does it end with exactly one next step? For any FAIL, show the smallest fix — don't rewrite the whole thing.

🧽 The de-slop pass

The last step before anything AI-drafted goes public: strip the tells that make writing smell like a robot.

Final pass on this draft: [paste]. Remove AI tells: stacked em dashes, "it's not X, it's Y" constructions, "delve," "game-changer," rule-of-three sentences, and any paragraph that says nothing. Keep my voice and my weird phrasings. Show me only the lines you changed.

Techniques adapted, in our own words, from Peter Yang's writing at creatoreconomy.so — the role/goal template, example-feeding, meta-prompting, decision docs, pass/fail evals, and his "no slop" editing pass.

Skills we learned from

Sabrina Ramonov — prompts like job descriptions

Sabrina Ramonov (sabrina.dev) teaches AI automation for creators and describes good prompting as writing job descriptions, not wishes. These six moves are her techniques, rebuilt in our words for the ADHD workflow.

📋 TCCA: Task, Context, Constraints, Ask

Her four-part scaffold. The genius part for ADHD brains is the last A — the AI asks you questions, so you don't have to think of everything up front.

TASK: [what you want, one sentence]. CONTEXT: I run [business], my customer is [who], this is for [why]. CONSTRAINTS: under 300 words, plain language, no fake enthusiasm. ASK: before you start, ask me clarifying questions one at a time until you're 95% confident you understand — then go.

🥊 The sparring partner

Flips the AI from yes-machine to challenger. The impulse-purchase firewall on our Tools page is this move wearing a disguise.

Be my sparring partner on this plan: [paste plan or idea]. Don't encourage me. Find my blind spots, my weak assumptions, and the step I'm secretly avoiding because it's boring. Rank the three biggest risks. Then — only then — tell me the strongest version of the idea.

⛓️ The chain, not the essay

She rarely writes one giant prompt — she runs short prompts in sequence, each building on the last answer. For ADHD brains: smaller turns, less wall-of-text.

We're going to work in a chain, not one big answer. Step 1 only: give me 5 options for [thing], one line each. Wait. When I pick one, I'll say NEXT and you'll go one level deeper on my pick. Never answer more than one step ahead.

🗣️ Clone your voice, once

Paste in your real writing once, save it as standing instructions, and every future draft sounds like you — not like a press release.

Learn my voice from these samples: [paste 3–5 things you actually wrote]. Build me a short style guide: 5 things I do, 5 things I never do, and a banned-words list. I'll save it to your custom instructions so we never do this again. Keep it under 150 words.

🔁 One thing becomes five

Her repurposing engine, sized down: make a thing once, let AI multiply it. The ADHD win: the creative burst gets spent once, not five times.

Here's one thing I made: [paste post / email / script]. Turn it into: 1) a LinkedIn post, 2) three short video hooks, 3) an email to my list, 4) a one-line X post, 5) a question that starts a conversation. Match my voice from our style guide. Label each clearly so I can grab-and-go.

💎 Mine the session for a skill

Her compounding move: before closing a chat that went well, extract the repeatable part. It's SNAP for workflows — save the game, keep the loot.

That worked. Before I go: look back at this whole conversation and turn what we did into a reusable recipe — the prompt, the steps, and the one mistake to avoid next time. Format it so I can paste it into a note called "[task] recipe" and run it cold next month.

Techniques adapted, in our own words, from Sabrina Ramonov's public posts at sabrina.dev and the Blotato blog — her TCCA structure, clarifying-questions gate, sparring-partner reframe, voice cloning, repurposing chains, and skill extraction. She also publicly champions voice dictation (her posts recommend the VoiceInk app on Mac for keeping voice data on-device) — which is the whole spirit of YAP.

Putting it together

Where each skill bolts onto the method

When you're…Method stepReach for
Full brain, can't sortYAPMaster prompt · voice dictation · TCCA's "ask me questions" gate
Sorted but can't chooseMAPThe chain (5 options, pick one) · decision copilot
Chosen but can't startTAPEmergency overwhelm · role + goal opener · body double
Done (or interrupted)SNAPSave card · mine-the-session-for-a-skill
About to publish or buyThe firewallSparring partner · pass/fail checker · de-slop pass

Skills are learnable. Come learn them live.

Every one of these gets taught hands-on in the live classes — with the room's phones out and the postcard rule on the wall.

See the classes → Get the free ebook