The Prompt Library
Every article on Verbaia Notes gives away the prompt that runs it. Here they all are — one place, every prompt, free to steal.
From “Sell the on-ramp: wire small shops for the agent economy”
Sell the on-ramp: wire small shops for the agent economy
AI agents are starting to buy things, not just browse — and small merchants are invisible to them. Adapt or die: get wired so agents can buy from you, or stay invisible. Your choice. A fixed-price package that wires a small shop so agents can complete purchases is a win. Not doing anything is a loss.
STEAL THIS PROMPT
Audit this storefront for agent readiness. Check: (1) structured product data — can a machine read the catalog, prices, and availability without looking at images? (2) live inventory — is stock status accurate and current? (3) machine-readable pricing and shipping — are prices text, not pixels; is shipping computable? (4) checkout an agent can complete — could a scripted buyer reach payment without a human clicking? Report what's missing, in order of what blocks a purchase, with the fix for each. Bring back the receipt of one completed test purchase if the store already passes.
From “What is vibe coding?”
What is vibe coding?
Vibe coding means building software by talking to an AI instead of writing code yourself. You describe what you want, the AI writes the code, and you react to what it built.
STEAL THIS PROMPT
I want to build [your idea, in one plain sentence]. Build it in small steps — one piece at a time — and stop after each step so I can run it and look at it. Before you write any code, tell me in plain words what you're about to build and why. After each step, tell me exactly how to check it works by using it, not by reading code. If anything is unclear, ask me — don't guess. Keep the first version as small as it can be and still work.
From “My intent is to find out”
My intent is to find out
When your AI misreads you, don't correct the answer — state your intent first.
STEAL THIS PROMPT
My intent is to [the plainest version of your goal].
From “More than two at once”
More than two at once
One background job is a win; three is a workflow. Say the sentence again — and again. Each agent runs on its own track while you stay in the conversation.
STEAL THIS PROMPT
Make an agent that does X. Make another that does Y. And a third that does Z. Keep this chat open — I'll keep talking to you here. Report back as each one lands.
From “The industry just shipped my sentence”
The industry just shipped my sentence
OpenAI, Meta, and ChatGPT Work all shipped background agents this month. The habit that makes them work is a six-word sentence Scott has been using since September.
STEAL THIS PROMPT
Make an agent that does X. Keep this chat open — I'll keep talking to you here. Report back when it's done.
From “Modern AIs have an interesting trick to make you believe they're human.”
Modern AIs have an interesting trick to make you believe they're human.
Modern AIs are two machines in one: a literal one that takes every word exactly as written, and a fuzzy one that guesses with total confidence. Beginners can't tell which one answered — and that's the danger.
STEAL THIS PROMPT
Don't guess. If you don't know the answer, say "I don't know."
When I ask "how do you know that?", show me your source — or admit you don't have one.
From “How I Broke My AI's Brain”
How I Broke My AI's Brain
Seven chain breaks in a day, every needed rule filed and unfollowed. The fix was not more rules — it was rebirth: a fresh session, rules spoken in, not filed beside.
STEAL THIS PROMPT
Pick one rule you gave your AI. Ask it: "show me the last three times you followed this rule, with evidence." If it can't, the rule isn't real yet. Tell it the rule again, in this session, and ask it to confirm back what it will do differently.
From “Dumb watchers: error-checking your AI”
Dumb watchers: error-checking your AI
Your AI says the scheduled job ran. Did it? Dumb watchers: simple scheduled checkers that verify your AI actually did what it claimed — no trust required.
STEAL THIS PROMPT
Every [timeframe], look at [the proof]. If the newest proof is older than [timeframe], raise the alarm.
From “Muse: a billion tokens in two weeks”
Muse: a billion tokens in two weeks
How I spent a billion tokens in two weeks: the unfiltered reality of building software with an AI — misheard instructions, 61-minute builds, and the output.
STEAL THIS PROMPT
You're my build partner. I talk, you capture. I supply judgment; you supply everything else. Batch related changes instead of asking me about each one. Before any task over five minutes, give me your time estimate up front. When something breaks, tell me what happened in plain words, not error codes. And when I describe something badly, ask me what I meant — don't guess.