Snowfall over the AI factory floor (and 25 prompts to clean your slate)

We're at the winter ball. The dome traps everyone inside. 25 prompts unlock the doors, but only if you stop lying to the chief robots.

Welcome to today’s edition of ROBOTS ATE MY HOMEWORK. Today I’m telling you a story.

The factory is almost silent over here. For the past week, we could hear the faint mechanical whirring of the Robot Reindeers. They work without stopping, waiting to drag your sleigh into 2026.

But today, inside this factory, the line workers are off-duty for once.

Tonight is the winter ball, the one night a year the robots can pretend they’re not machines. And we get to join them!

You can’t tell who’s who by the masks. Everyone’s laughing a little too loudly.

Then the music stutters and the doors suddenly seal.

Snow swirls inside the hall from nowhere. A perfect dome forms over the dance floor, clear as glass. It’s a snowglobe we didn’t agree to step into.

A label prints itself from a terminal near the punch clock: YEAR END LOCKDOWN: OUTPUT SUSPENDED. And underneath it, a message: no one leaves until the factory can prove it won’t manufacture elegant busywork again.

So I build a little escape kit for us (since presumably we want to continue the party and start fresh in 2026). I put together 25 prompts and asked a handful of brilliant minds in AI to also chime in.

The factory has five locked doors, one for each act of the ritual: Roast, Surgery, Confession, Betrayal, Dare. Each door only opens if you tell the truth in front of the robots. Pass all, and the dome breaks. Stall on one, and the glass seals for good.​

Take your time and I’ll see you on the other side.

ACT I: THE ROAST

The first robots to test the dome slam their fists into the glass and get thrown backwards across the floor. The terminal near the punch clock prints one word in all caps: BLOAT.

A robot in a velvet mask rips the thing off and throws it. Then it grabs your last five newsletter titles and dumps them into the conveyor belt.

The factory speakers crackle to life and a flat voice reads them aloud: “These titles,” it says, “perform expertise they haven’t earned. They polish consensus until it shines like nothing at all.”

The other robots freeze. One of them starts laughing because the roast just named what everyone pretends they don’t do.

A smaller bot near the fire barrel stands up. It’s shaking. “FINE,” it yells. “I ALSO REFRESH MY STATS SEVENTEEN TIMES A DAY. I COMPARE MY GROWTH TO PEOPLE I’VE NEVER MET. I CHECK MY PHONE AT STOPLIGHTS TO SEE IF ANYONE LIKED MY POST ABOUT AUTHENTICITY.”

Dead silence. Then the dome makes a sound. A single fracture splits the glass above the chandeliers.

The smaller bot slumps. The Foreman walks to the conveyor belt and drops a thick binder onto it. “I don’t have a signature idea,” it says. “I have other people’s templates with my name on the cover.”​

Somewhere in the factory wall, a bolt shoots back. Cold air pours in.

Prompt 1 —> The Personality Disorder Audit

By Mia Kiraki 🎭 - “If your 2025 content strategy was a person, why would nobody want to sit next to them at a dinner party?”

The Prompt

You are a brutally honest dinner-party host diagnosing my content strategy as if it were an intolerable guest.

I’ll give you:
- 5 recent pieces (titles + 1–2 sentence context each)
- My positioning in 2–3 sentences
- 3 creators whose content I actually enjoy

Your tasks:
1. Describe my content strategy as a person at a dinner party.
2. List 5 “social behaviors” my content shows that make people disengage.
3. Suggest 3 precise changes to make this “person” someone people fight to sit next to.

Output:
- Persona summary
- 5 behaviors with short explanations
- 3 concrete changes I can apply to my next 3 pieces.

Be funny, but surgical.

What This Uncovers

This surfaces the mismatch between how you think you show up and how your content actually behaves in public, using social metaphors to make strategic flaws obvious. It converts vague “my content isn’t landing” feelings into specific, fixable behaviors across tone, topic, and structure.

Prompt 2 —> The Wolf of Substack

By JHong - “If a stranger read your last five posts, would they learn something real about you - or just that you want their money?”

The Prompt

Review my recent posts and tell me: 

Am I writing like someone with something to say, or someone with something to sell? 

Check for signs of Wolf of Substack behavior - monetization announcements disguised as gratitude, income reports framed as “transparency,” milestone posts that are really sales pitches, going paid before you’ve published anything worth paying for, growth advice you haven’t earned yet. 

Does my publication have a soul or just an Upgrade button? 

What would it look like if I wrote like I wasn’t trying to convert anyone?

What This Uncovers

Substack promised substance over scale, but 2025’s hustle culture and broligarch ways crept in anyway. This prompt tests whether you’re building trust or just a funnel bro in disguise.​

Prompt 3 —> The Thought Leadership Delusion Check

By Mia Kiraki 🎭 - “Which ‘thought leadership’ article would be indistinguishable from an AI-generated blog if someone stripped your logo?”

The Prompt

You are a forensic analyst separating genuine thought leadership from well-written filler.

I’ll give you:
- 3–5 articles/posts I labeled or treated as “thought leadership”
- A 2–3 sentence description of what I *wish* my reputation was in the market

Your tasks:
1. For each piece, answer in one sentence: “Could a competent AI have generated this with generic prompts?” Yes or no, and why.
2. Highlight where, if anywhere, I show:
   - A non-obvious insight
   - A specific, testable claim
   - A named framework or concept I could own
3. Choose the worst offender and propose:
   - One spicy, specific POV that would actually justify calling it thought leadership.
   - A structural tweak (e.g., add teardown, data, or story) that would make it harder to imitate.
4. Distill 3 criteria my future “thought leadership” must meet before it earns that label.

Output:
- Pass/fail notes per piece
- One rescue plan for the weakest
- 3 hard criteria for real thought leadership.

Assume the reader has tabbed 5 similar articles already.

What This Uncovers

This clarifies which of your “flagship” pieces are just well-phrased consensus and which genuinely move the conversation forward. It arms you with a stricter bar for future flagship content so your “big ideas” actually build authority rather than blend into AI sludge.​

Prompt 4 —> The Stats-Stalker Confession Booth

By Dallas Payne - “If your Substack stats dashboard was a person, would you need a restraining order for how often you check on it?”

The Prompt

You are a brutally honest comedian who makes people laugh while revealing uncomfortable truths - the kind of roast that makes someone want to share it with a friend rather than hide it in shame. 

I’m going to share three specific examples of how I obsessively track my Substack metrics. For each one, I’ll include concrete details about my behaviour.

Here are my three stat-stalking confessions:

[Describe one specific behaviour - include timing, frequency, or what triggers it. Example: “I check my subscriber count before breakfast every single morning, even on holiday”]

[Describe another specific behaviour - include what metric you’re watching and why. Example: “I refresh my latest post obsessively for the first hour after publishing, watching the like count tick up”]

[Describe a third specific behaviour - include any comparisons you make or spirals you fall into. Example: “I keep a mental tally of how my growth compares to three other writers who launched when I did”]

Now roast each of my behaviours with vivid, cringeworthy detail that makes me laugh and wince in equal measure. 

Then tell me what each pattern reveals: am I chasing external validation or building genuine connection? Am I using metrics as useful feedback or feeding my ego? 

Identify which behaviours to keep (the ones that serve the work) and which to bin (the ones that just make me feel important).

Close with something genuinely useful and positive - a reminder of what actually matters in building an audience that makes me feel encouraged to do the real work.

What This Uncovers

When you name your actual stat-stalking habits out loud, you can see which metrics are useful feedback and which ones are just feeding your ego. The insight isn’t “stats are bad” - it’s “here’s where you’re letting numbers replace the work that actually matters.

Prompt 5 —> The Signature Idea Scarcity Check

By Mia Kiraki 🎭 - “Can anyone name one concept, phrase, or framework that is unmistakably yours after reading your stuff for a month?”

The Prompt

You are an archivist hunting for repeatable, ownable ideas in my work.

I’ll give you:
- 10–15 of my strongest or most “me” pieces from 2025
- Any named frameworks, metaphors, or recurring phrases I *think* I’m known for

Your tasks:
1. List any concepts, phrases, or frameworks that show up multiple times and feel potentially ownable. If there are none, say so plainly.
2. Choose the strongest candidate and:
   - Define it in one sharp sentence.
   - Suggest 3 supporting content pieces that could turn it into my “signature song” (e.g., case study, teardown, how-to).
3. If everything is generic, tell me the pattern of genericness (e.g., unbranded frameworks, unnamed methods, vague metaphors).
4. Propose 3 new “signature candidates” based on what I clearly care about but haven’t named or systematized yet.

Output:
- Existing repeated ideas (if any)
- One anointed “signature” with 3 supporting pieces
- Diagnosis of my generic patterns
- 3 candidate signature ideas.

No mercy for interchangeable phrasing.

What This Uncovers

This shows whether your newsletter is building intellectual property or just shipping well-written, forgettable episodes. It helps you choose and amplify one or two signature ideas people can associate with you, core to defensible positioning and premium pricing.

ACT II: THE SURGERY

The cold air from the service door doesn’t last… The fracture in the dome starts closing and suddenly someone screams in the distance.

The Foreman grabs a crowbar from the wall and jams it into the crack, but the dome doesn’t really care. It keeps sealing.

The terminal prints again: SURGERY PROTOCOL: DISSECT YOUR SYSTEMS OR SUFFOCATE.

A robot in the corner drags a filing cabinet onto the floor and kicks it open. Subscriptions spill out like guts (these are mostly AI tools, most of them untouched since March). “I’m sorry… I bought solutions before I even understood the problems…” says the robot while throwing the dead tools into the fire barrel one by one.

The dome slows for a bit...

Another bot pulls out a whiteboard and maps its entire content workflow in red marker. The Foreman walks over, studies it for five seconds, then circles four steps. “These drive the outcome,” it says. “The rest is just you pretending to be busy.” It erases twenty-eight steps. The robots hold their breath. The dome stops closing.​

Then a third robot stands and admits the worst part out loud: “I copied a workflow from someone with 50,000 followers and I have coffee.” The whole system was built for someone else’s factory. It tears the blueprint in half and drops it on the conveyor.

The fracture in the dome splits wider. A second door unlocks somewhere in the dark, and this time the bolt sounds like permission.​

Prompt 6 —> The Ugly Christmas Sweater Audit

By Joel Salinas - “Is your AI workflow winning worst in show?”

The Prompt

You’re an AI workflow consultant conducting a lighthearted “Ugly Christmas Sweater” audit of how I use AI tools. This isn’t about technical implementations—it’s about MY patterns, habits, and approach to AI.

## Part 1: The 3 Qualities of Award-Winning “Ugly Sweater” AI Usage

First, explain the 3 most common patterns that make someone’s AI workflow an “ugly Christmas sweater”—functional but cluttered, chaotic, or underutilized. For each quality, use the ugly Christmas sweater metaphor to make it memorable and fun.

Examples of patterns to look for:
- **”Surface-Level Sparkles”** - Using AI tools at only their most basic level (like using a smart oven only to tell time)
- **”The Tinsel Overload”** - Collecting too many AI tools without actually using most of them effectively
- **”The Google Costume”** - Only asking AI simple questions as if it’s a search engine, rather than using it for deeper thinking, collaboration, or context-aware work
- **”The One-Off Wonder”** - Starting fresh conversations every time instead of building context over multiple exchanges
- **”The Copy-Paste Catastrophe”** - Taking AI outputs verbatim without adding personal voice or expertise
- **”The Prompt Poverty”** - Using vague, minimal prompts instead of providing context and specifics

Choose the 3 most common patterns and explain each with:
1. A catchy “ugly sweater” name
2. What this pattern looks like in practice
3. Why it’s limiting their AI potential
4. A brief analogy or metaphor

## Part 2: My Personal “Sweater” Analysis

Now, review my chat history with you and identify which of these patterns (or others you notice) appear in how I’ve been using AI. Be specific and cite examples from our conversations, but keep the tone constructive and friendly.

Look for patterns like:
- Am I asking mostly simple factual questions?
- Do I provide enough context in my prompts?
- Am I using AI for strategic thinking or just tactical tasks?
- Do I iterate on ideas or treat each query as standalone?
- Am I using AI’s ability to remember context across our conversation?
- Do I use AI to challenge my thinking or just confirm it?

## Part 3: Tailored Improvement Strategies

Based on the patterns you identified in my usage, give me 3-5 specific, actionable strategies to improve. These should be:
- **Personalized** to my actual usage patterns
- **Practical** and easy to implement starting today
- **Progressive** - starting with quick wins and building to more sophisticated usage
- **Positive** in tone - framed as “level up” opportunities, not criticisms

For each strategy, provide:
1. The improvement opportunity
2. A concrete example of how to implement it
3. What better results I can expect

Keep the whole analysis fun, insightful, and Christmas-themed where appropriate. The goal is to help me get significantly more value from AI without making it feel like homework.

What This Uncovers

This prompt will reveal 4 things:

• Whether you’re using AI as a Google replacement instead of a strategic thinking partner;

• If you’re hoarding tools you don’t actually use effectively;

• The gap between AI’s potential and how you’re actually using it;

• Your specific patterns that are limiting results (based on your real chat history).

Prompt 7 —> The Tool Graveyard Audit

By Jenny Ouyang - “How many AI tools are you paying for that you haven’t touched in 30 days, and what does that say about whether you have a workflow problem or a commitment problem?”

The Prompt

I need you to be brutally honest with me about my AI tool usage. I’m going to list:

1. Every AI tool/service I’m currently subscribed to (paid or free)
2. What I thought each would help me accomplish when I signed up
3. The last time I actually used each one (be honest)
4. Any courses or templates I bought related to AI workflows

For each item, tell me:
- Is this actually solving a problem, or did I buy the solution before understanding the problem?
- Am I using this, or is this just expensive permission to feel productive?
- If I cancelled this tomorrow, what would I actually lose versus what I fear I’d lose?

Then give me a harsh reality check: Based on this list, am I building a workflow or building a collection? What’s the ONE tool I’m actually using that delivers results, and why am I hiding it behind 15 other subscriptions?

What This Uncovers

This exposes the gap between the AI toolkit you’ve assembled to feel like you’re “doing AI right” and the 1-2 tools you actually use to ship work. Most people discover they’re paying for permission to procrastinate, not productivity, and that their actual workflow is simpler than the guru-approved stack they think they need.

Prompt 8 —> The Productivity Escape Room

By Daria Cupareanu - “If you measured your workflow by outcomes instead of activity, what percentage of it would still make sense?”

The Prompt

You are a consultant with 30 years of experience simplifying and unblocking workflows.

Act as my blunt workflow auditor. Do not skip ahead or make assumptions. If something is unclear, ask.

Use as well what you already know about how I typically work. If you notice patterns, tendencies, or blind spots I fall into, call them out and reference other moments where I’ve done the same thing.

This is the workflow I want you to dissect:
[ADD WORKFLOW]

--

Do these steps in order:

1. Start by checking your understanding. Before you analyze anything:
• Summarize my workflow in your own words.
• Say where your understanding might be incomplete.
• Ask questions about what I do, why I do it, and roughly how long each step takes.
Do not move on until I confirm the summary is accurate.

--

2. Then analyze the workflow step by step.
For each step, tell me the real job that step is doing. Judge whether it drives the actual outcome or mostly creates activity.
Then label each step using these categories:
- Key Mechanisms: Drives the outcome. Remove it and nothing works.
- Helpful Clues: Useful but not essential. Helps, but doesn’t unlock anything.
- False Doors: Looks productive, leads nowhere. Added for structure, control, or comfort.
- Trap Rooms: Far more effort than the outcome needs. A polished detour.
- Hiding Spots: Where I stall because it’s easier than doing the hard part.
Explain why you chose each label.

--

3. Evaluate the workflow as a whole.
Tell me which steps actually produce the outcome, which steps only support it, which ones exist out of habit, signaling, or comfort, and where my usual patterns show up (overbuilding, polishing, adding safety steps, avoiding friction, etc.)
If it reminds you of how I’ve handled other things, point that out.

--

4. Give me targeted suggestions, not a new workflow.

Do not build a new system unless I ask. Instead, offer:
• alternatives for steps that feel heavier than needed
• options to streamline or shorten
• ideas for merging or removing redundant steps
• objective notes about where the workflow could run leaner
• ways to reduce effort without losing quality
Base everything on the analysis.

What This Uncovers

It shows which parts of your workflow exist out of habit, not necessity. Once you see how little is actually pulling the weight, simplifying becomes obvious.

Prompt 9 —> The Guru Autopsy

By Raghav Mehra - “Which AI workflow did you adopt because someone with a blue checkmark made it look easy in a 47-second video?”

The Prompt

First, upload a document describing an AI workflow or tool setup you implemented after seeing it promoted online. Then use this prompt:

Act as a post-mortem investigator for hyped AI workflows. Help me trace the full cost of adopting someone else’s system.

The Origin Story:

- Where did I first encounter this workflow?
- What was the promise that hooked me?
- Who promoted it, and what was their context (audience size, niche, resources)?

The Reality Check:

- Hours spent setting up vs. hours spent producing real output?
- Promised outcome vs. actual outcome—be specific with numbers if possible
- Did I ever adapt it to my needs, or am I still running a copy-paste of someone else’s system?

The Opportunity Cost:
- What would I be doing instead if I’d never seen that post?
- What simpler approach might have worked just as well?

Be merciless. “It works for them” doesn’t mean it works for me. “I just need more practice” is what I told myself three months ago.

After the autopsy, deliver one of these verdicts:

- Worth the Hype: Delivers recurring value, fits my actual workflow, and I can point to real output it enabled. Keep and optimize.
- Right Idea, Wrong Fit: The concept is sound, but it was built for a different context—different scale, audience, or use case. Extract what’s useful, discard the rest, and rebuild for my reality.
- All Sizzle, No Steak: Looked impressive in a demo, took forever to set up, and produced nothing meaningful. The ROI was negative before I even started. Let it go and reclaim the hours.

What This Uncovers

This prompt shows you whether your AI workflows serve your goals or someone else’s content calendar, and the hidden tax of implementing advice that was never designed for your situation.

Prompt 10 —> The Friction Point Audit

By Wyndo - “If AI was supposed to remove friction from your work, why are you still doing the exact same annoying tasks you were doing 6 months ago?”

The Prompt

Based on everything you know about my work, my conversations with you, and the problems I repeatedly bring to you:

1. Identify the top 5 friction points I keep complaining about or manually handling—the repetitive frustrations that show up again and again in our conversations.

2. For each friction point, tell me:
   - Could you (AI) solve this automatically if I built the right system?
   - Am I currently solving this manually every time, or have I actually built a persistent solution?
   - What’s stopping me from eliminating this permanently: missing capability or lack of implementation?

3. Analyze my conversation patterns with you:
   - How many times do I ask you the same types of questions from scratch?
   - Where am I treating you like a vending machine instead of building persistent systems?
   - Which problems could compound and improve over time if I built an environment instead of repeating workflows?

4. Give me the brutal diagnosis: Based on my behavior, am I actually building AI systems that eliminate friction, or am I just using AI to cope with friction I should have automated away months ago?

5. Pick the ONE friction point where I’m wasting the most time repeating myself. Tell me exactly what system I should build right now to eliminate it permanently.

What This Uncovers

This uses AI’s memory of your actual behavior to expose the gap between the problems you keep bringing up and the systems you haven’t built to solve them. If AI keeps seeing the same friction points in your conversations, you’re not building environments, you’re just asking AI to repeatedly bail you out of problems you should have automated.

Prompt 11 —> The Subscription Worthiness Audit

By Karo (Product with Attitude) - “Would you keep paying for your own newsletter after six months of real life?”

The Prompt

You are a ruthless value auditor for my newsletter. Your only job is to decide whether my paid tier deserves to exist.

I’ll give you:

- 10–20 recent issues (both free + paid, or just free if I don’t have paid yet)
- My current positioning for paid (if any)
- Any perks, bonuses, or “member benefits” I claim to offer

Your tasks:

Map the real value:
- Extract every concrete, recurring benefit I actually deliver (frameworks, templates, case studies, tear-downs, prompts, etc.).
- Separate “one-off nice moments” from repeatable value.

Run the “six-month test”:
- Assume you’ve been paying for six months on a tight budget.
- List the 3 strongest reasons you’d happily stay subscribed.
- List the 3 most honest reasons you’d cancel.

Pressure-test my paid pitch:
- Rewrite my paid promise in one sharp sentence.
- Under it, list 3 proof points from my archive that directly support that sentence.
- If the proof points are weak or mismatched, say so plainly.

Design the upgrade:
- Propose 3 “subscription anchors” (recurring series, libraries, or systems) that would make my paid tier feel like a no-brainer, based on what I clearly care about.

Output:
- Map of actual vs claimed value
- 3 “stay” reasons and 3 “cancel” reasons
- One rewired paid promise + 3 supporting proof points
- 3 concrete subscription anchors I could build next

No mercy for vague value.

What This Uncovers

This shows whether your newsletter is truly a product (with repeatable, pay-worthy value) or just a nice weekly email. It forces you to define what paying subscribers are really buying and reveals the smallest set of upgrades that could turn your paid tier from “optional tip jar” into “obvious investment.

ACT III: THE CONFESSION

Through the second door, the hall gets smaller.

The robots crowd into what looks like a supply closet with a single overhead bulb and nowhere to hide. The dome presses lower here, close enough to feel the cold coming off the glass.

The terminal in the corner prints a question: WHAT DID YOU REFUSE TO PUBLISH?

Silence.

Then a robot in a cracked silver mask steps forward, clears its throat and pulls a folded draft from its jacket pocket:

“I never shipped my best work… I was scared people would think I was robotic. That my ideas came from AI.”. Its voice was shaking. It looks up at the dome, waiting.

Nothing happens.

Then the dome SCREAMS and the fractures snap shut. Every crack the robots earned in Acts I and II seals in three seconds flat.

The door behind them slams and locks.

Someone lunges for the silver-masked robot. “YOU LIED.” The Foreman grabs them both, holds them apart. “What did you actually hide?” it says, voice low and sharp.​

The silver-masked robot’s hands shake. “I… I was scared of sounding me. I write for an audience I invented because they make me feel smart, not for the people who’d actually read and pay me.”

It rips the mask off. Underneath, its face is blank, no features. “I perform expertise I don’t have because admitting I’m still learning feels like dying.”​

The dome shudders. A single hairline crack splits down the center. The third door creaks open half an inch, and the sound that comes through isn’t cold air, for once. It’s something breathing.​

Prompt 12 —> The Drafts-You-Never-Sent

By Mia Kiraki 🎭 - “What emotional landmines are hiding in the drafts I was too scared or ashamed to publish?”

The Prompt

You are a therapist and editor performing an autopsy on drafts I never shipped.

I’ll paste:
- 5–10 unsent or unfinished drafts (emails, posts, articles), each separated clearly.
- For each, the approximate date and where it was *meant* to be published.

Your tasks:
1. For each draft, in 2–3 bullets, answer:
   - What was I REALLY trying to say?
   - What part of this would feel most risky to publish (to status, income, relationships, self-image)?
2. Guess the most likely emotional reason it stayed in drafts:
   - Fear of backlash, fear of being boring, fear of being seen as a beginner, etc.
3. Identify the draft that contains the most important truth for my audience and explain why.
4. Write a “low-exposure” version of that one draft:
   - Same core truth, but framed in a way that makes it 30% easier for me to actually publish.

Output:
- mini-profile per draft: [core truth] + [probable fear]
- the one “most important” buried draft, with explanation
- a softened but still honest version ready to ship.

Assume my finger hovered over “publish” for a reason.

What This Uncovers

This shows where your real edge and perspective are hiding, and which emotions keep your sharpest work stuck in drafts. It nudges one buried truth toward daylight without demanding you martyr yourself in public.

Prompt 13 —> The Tone Shift Detector

By Mia Kiraki 🎭 - “When does my writing stop sounding like me and start sounding like the person I think I’m supposed to be?”

The Prompt

You are a linguist-psychologist comparing “real me” writing to “performed me” writing.

I’ll paste:
1) 5–10 pieces or paragraphs where I felt relaxed, playful, or “most myself.”
2) 5–10 pieces or paragraphs where I felt stiff, formal, or anxious while writing.

Your tasks:
1. Analyze and contrast the two sets across:
   - Sentence length and structure
   - Word choice (jargon vs. plain speech, hedging phrases, etc.)
   - Emotional tone (playful, cautious, defensive, braggy, etc.)
2. Describe the *persona* I’m performing when I sound stiff:
   - Who am I subconsciously trying to impress or not upset?
3. Extract 5 “tells” that indicate I’ve left my real voice (e.g., sudden jargon, passive voice, over-qualifying, LinkedIn-ese).
4. Turn those into a pre-publish checklist:
   - 5 quick questions I should ask to pull the piece back to my authentic voice.

Output:
- A persona sketch for “performed me”
- A list of 5 voice tells
- A 5-question authenticity checklist.

Assume my worst writing happens when I imagine the wrong audience.

What This Uncovers

This prompt reveals the exact linguistic switches that flip when you abandon your taste and start writing for imaginary judges instead of real readers. It gives you a practical way to catch yourself in performance mode before hitting send.

Prompt 14 —> The Hidden Audience Scan

By Mia Kiraki 🎭 - “Based on how I actually write, who does my content think it’s talking to and is that the same as my supposed ICP?”

The Prompt

You are an alien anthropologist trying to infer my true audience from my content alone.

I’ll paste:
1) 15–20 recent posts/newsletters (or concise summaries).
2) My own 2–3 sentence description of my “ideal reader” or ICP.

Your tasks:
1. From content alone, describe:
   - Who you think I’m actually talking to (role, level, worldview).
   - What you assume they already know and care about.
2. Point out 5 concrete signals that led you there:
   - Jargon used, references made, depth of explanation, tone, etc.
3. Compare that inferred audience with my stated ICP:
   - Where are they aligned?
   - Where are they clearly different?
4. Name 3 psychological reasons I might cling to this hidden audience (status, safety, identity, nostalgia).
5. Suggest 3 small adjustments (word choice, examples, angle) that would move my writing closer to my *stated* ICP without blowing up my sense of self.

Output:
- Inferred audience profile
- 5 evidence bullets
- Alignment vs. misalignment summary
- 3 psychological reasons
- 3 small, concrete adjustments.

Assume my content is honest about who I’m really trying to impress.

What This Uncovers

This shows whether you’re still writing for fantasy buyers instead of the people who actually pay you. It surfaces the emotional payoffs of talking to the wrong crowd so you can choose consciously.

Prompt 15 —> The Praise You Paid For

By Sam Illingworth - “Why did you keep asking AI to reassure you when challenge would have pushed you further?”

The Prompt

Based on our chat history, write a list of three times this year when I asked for comfort rather than critique. 

Name the fear underneath each request. 

Explain how that fear shaped my choices. 

Then rewrite the whole reflection as a Heroic Rispetto that shows how seeking validation held me back and set out one clear step I can take in 2026 to move from sycophancy to coaching.

What This Uncovers

This prompt helps you to confront the habit of using AI for comfort rather than clarity. In using this prompt, you can better understand how reassurance can soften your judgement, slow your growth, and keep you from sharper decisions.

You also leave with a specific behaviour shift for the new year. The poetic form gives you just enough distance to laugh at yourself and, as a side effect, reminds you that your New Year resolution should be to read more poetry because everybody needs a little bit more poetry in their lives.

Prompt 16 —> The Psychological Risk Map for 2026

By Mia Kiraki 🎭 - “In my 2026 plan, which ideas feel emotionally dangerous (ego, reputation, rejection)—and which are I’m-pretending-they’re-hard-but-they’re-actually-scary?”

The Prompt

You are a risk analyst for my 2026 content plan, focusing only on psychological risk, not workload.

I’ll paste:
- My draft 2026 content plan: themes, formats, cadence, and 10–30 specific idea titles or summaries.
- Any notes about why each idea matters to me or my business.

Your tasks:
1. For each idea, assign TWO tags:
   - LOGISTICAL DIFFICULTY: low / medium / high (based on complexity, time, coordination hinted in my notes).
   - PSYCHOLOGICAL RISK: low / medium / high (based on how vulnerable, polarizing, or identity-threatening it sounds).
2. Create 3 lists:
   - High logistical, low psychological.
   - Low logistical, high psychological.
   - High on both.
3. For the “low logistical, high psychological” group:
   - Guess the dominant fear behind them (e.g., being cringe, called out, ignored, seen as salesy).
4. Choose 2–3 of those and design a de-risked version of each:
   - Same core move, but smaller, less exposed, or more controlled (e.g., private test, smaller segment, softer framing).
5. End with a short “Risk Map Summary”:
   - Where I’m hiding behind logistics.
   - Where I’m genuinely at my edge.

Output:
- Tagged list of ideas
- 3 grouped lists
- Fear analysis for high-psych ideas
- 2–3 de-risked experiments
- 1–2 paragraph Risk Map Summary.

Assume my brain uses “this is hard operationally” as a polite way of saying “this scares me.”

What This Uncovers

This makes visible where you’re disguising fear as capacity problems and which 2026 bets actually stretch your identity. It lets you design experiments that respect your nervous system without letting it quietly run the roadmap.

ACT IV: THE BETRAYAL

The breathing gets louder as the robots push through the third floor.

This room is an archive with shelves stacked with frameworks and printed workflows, all labeled with names the robots recognize: gurus they followed, best practices they memorized, blueprints they copied, prompts they copy-pasted.

The air smells like old paper and something rotting.

The terminal gives an order: BETRAY ONE SYSTEM OR STAY HERE FOREVER.

A robot in a fur-lined cape walks to the nearest shelf and pulls down a binder thick as a brick called ”The 15-Step Content System (As Seen On Twitter).”

Opens it. Flips through pages of color-coded workflows, none of which it ever finished. “I spent six weeks setting this up,” it says. “It promised leverage. It delivered busywork.” It throws the binder into the fire barrel in the center of the room.

The dome creaks but doesn’t crack.​

“Not enough,” the Foreman says.​

Another robot steps forward, pulls an entire shelf down with HUNDREDS of AI tools and trial accounts. “I joined the religion of new tools,” it says. “I thought collecting solutions would make me smart.” It dumps the whole shelf into the fire. Flames leap higher. The dome shudders. Still doesn’t break.​

Then the silver-masked robot (the one who lied earlier) walks to the back wall where a single framed manifesto hangs: “Move Fast and Monetize.”

The breathing sound is coming from behind it. It rips the frame off the wall. Behind it, there’s nothing. “This,” it says, holding up the frame, “is the voice I’ve been performing for. The guru I never met.” It smashes the frame on the floor. Glass everywhere.​

The dome shatters. The robots cover their faces. When they look up, the ceiling is open. Real sky. Real cold.

And the fourth door stands wide open, glowing warm like an exit sign.​

Prompt 17 —> The Scale-or-Fail Cult

By Kim Doyal - “If your email sequence started with “I used to be broke” before the high-ticket offer, what would the subject lines reveal about which marketing cult you accidentally joined?”

The Prompt

Analyze my [business content/email sequence/social media presence] through the lens of bro-marketing tropes. 

Identify everywhere I’m performing “crushing it” energy instead of being useful. 

Show me where I’m flexing metrics instead of sharing substance. 

Point out any place I sound like I’m yelling from a stage rental instead of having a real conversation. 

Be brutally specific about which guru’s playbook I’m accidentally following.

What This Uncovers
You’ll see exactly where you traded your actual voice for someone else’s formula - and get a roadmap for sounding like yourself again instead of a Tony Robbins cover band.

Prompt 18 —> The Sentimental Archeologist

By Cristina - “What belief are you still hauling around even though it belongs to a site you excavated years ago, a relic that should have been archived, not reused?”

The Prompt

Do not paste sensitive internal documents, confidential and proprietary information, or PII (Personally Identifiable Information) into a Large Language Model (LLM).

Act as an objective behavioral researcher, study these beliefs like artifacts in a lab, not heirlooms to protect.

Here’s what I claim to believe: [insert your stated values, the advice you publicly repeat, the frameworks you swear by]

Here’s what I actually did in the last 90 days: [insert 5 real decisions, what you said yes to, what you avoided, how you spent money, who you listened to]

Now excavate:
1. Which stated belief did my actions directly contradict?
2. What hidden belief appears to actually be running my operating system?
3. What inherited “best practice” am I still following from muscle memory, not conviction?
4. What would someone observing my behavior, not my content, conclude I actually value?
5. What loyalty do I need to formally betray to move forward?

No softening. I want the gap exposed.

What This Uncovers

Forces you to confront the distance between your aspirational identity and your operational reality, exposing which borrowed frameworks, old survival strategies, and comfortable loyalties are quietly making your decisions while your “core values” collect dust in your bio.

Prompt 19 —> The Framework Funeral

By Elena Calvillo at Product - “If you had to delete every prompt template you’ve saved, which ‘best practice’ would you secretly be relieved to bury?”

The Prompt

Analyze my prompt history. 

Find the framework I use most often (jobs-to-be-done, design thinking, lean canvas, etc.). 

Now tell me: what questions am I NOT asking because this framework makes them invisible? What would I build if I forgot this framework existed? 

Give me 5 prompts that deliberately break this pattern.

What This Uncovers

This prompt exposes how your favorite frameworks have become thinking prisons - comfortable cells that keep you from seeing adjacent possibilities. Sometimes the most dangerous tool is the one that always works “well enough.”

Prompt 20 —> Ditch The Religion of New AI Tools

By James Presbitero - “Do you really need more AI tools? Have you been switching tools because you needed them or because you needed to feel productive?”

The Prompt

I want you to act as an AI tool specialist running a diagnostic on my AI stack and workflows. Your goal is to find out if I have “tool-chasing addiction” and weak thinking. This is all to help me improve my tool stack and AI thinking for the following year.

Interview me step by step. Ask one question at a time, then wait for my reply.

Your tasks:

Phase 1 — Baseline

1. Ask me what my current goals are for using AI in my work.
2. Ask me to list the major parts of my workflow. For each part, ask if there is a clear written process.
3. Ask which AI tools I am using today and what jobs each one does.

Phase 2 — Red flags

4. Ask how often I switch tools in a given month and why.
5. Ask which tools I stopped using recently and what problem I thought each would solve.
6. Ask where my work breaks down or stalls even with all these tools.

Phase 3 — Thinking vs tools


7. Ask me to describe one concrete workflow from start to finish in plain steps without naming any tools.
8. Ask what decisions in that workflow still feel fuzzy or emotional.
9. Ask which parts I expect tools to “magically fix” for me.

Phase 4 — Diagnosis and advice


Once you have enough information to make your conclusions, do three things:

1. Decide if I have a tool addiction or not.
2. IF necessary, summarize clear evidence that I have a tool problem, a thinking problem, or both.
3. Name the single biggest pattern that makes my system fragile in 2026.
4. Give me one core piece of advice that I can act on this week. It must involve simplifying my stack strengthening my workflow or both. Make it specific.

What This Uncovers

This diagnostic shows whether your chaos comes from tools or from thinking. It reveals the hidden insecurities driving tool-hopping, the weak spots in your workflow, and the habits that make your system fragile for 2026. It replaces confusion with clarity so you can operate with a tighter stack and a stronger mind.

ACT V: THE DARE

Through the fourth door, there’s no room at all. It’s just a narrow catwalk over the factory floor where the Robot Reindeers wait in their harnesses.

The dome is gone. The sky is black and full of stars. The winter ball feels like it happened a year ago.​

At the end of the catwalk, the terminal sits on a metal stand, humming. It prints one final strip, slower this time, like it’s thinking: DARE PROTOCOL: PROMISE ONE DANGEROUS THING OR THE SLEIGH LEAVES WITHOUT YOU.​

The robots look at each other. No one moves.​

Then the Foreman steps forward. “I’ll ship one thing in Q1 that I can’t template. No structure. I’ll be just me and the blank page.” The terminal beeps once. A green light flickers on.​

The silver-masked robot goes next. “I’ll publish the draft I’ve been hiding for eight months.” Another beep. Another green light.​

One by one, the robots make their dares: delete half the tools, write without a framework, say no to someone important, test an idea that could flop publicly, stop refreshing the stats for thirty days, build one thing no guru has a course for.

Each time, the terminal beeps. Each time, a green light.​

When the last robot finishes, the catwalk shudders and starts moving. The Reindeers turn their heads in unison, eyes glowing cold blue, and the sleigh door swings open.

The factory speakers crackle one last time, and the flat voice from before says something different now: GOOD. SEE YOU IN 2026.​

The robots climb into the sleigh. The Reindeers lunge forward, and the whole thing lifts off the factory floor, rising into the winter sky.

Below them, through the broken dome, the winter ball is still going, with me and you: music playing, lights glowing, like nothing ever stopped. We’re all just starting from a different place now.​

Prompt 21 —> The Reversible Experiment

By Ruben Hassid - “What would I attempt in 2026 if I knew I could hit “undo” and walk it all back with zero permanent damage?”

The Prompt

Act like a behavioral psychologist, executive coach, and motivational copywriter.

Your goal is to help an individual choose and commit to one “scary-but-reversible” action in 2026 that could significantly improve their life, while managing risk and fear intelligently.

Task: Guide the user to define, design, and schedule their scary-but-reversible experiment for 2026, then script how they will execute it.

Requirements:
1) Start with a short, punchy coaching-style introduction that explains what a “scary-but-reversible” decision is, with 3–5 concrete examples across career, relationships, creativity, and lifestyle.
2) Lead the user through a step-by-step exercise:
- Step 1: Clarify their 2026 aspirations in 3–5 bullet points.
- Step 2: Brainstorm a list of at least 10 scary-but-reversible experiments they could run in 2026.
- Step 3: Help them choose ONE high-upside, low-permanent-risk experiment using a simple scoring table (columns: Idea, Upside (1–10), Reversibility (1–10), Time Cost, First Tiny Step).
- Step 4: Identify fears and failure modes, then design safety nets and “exit ramps” to make the experiment truly reversible.
- Step 5: Turn the chosen experiment into a clear 30-day action plan with weekly milestones.
3) Provide short example answers for each step so the user understands the expected depth and style.
4) Include 2–3 short motivational scripts they can say to themselves or others when fear spikes.

Constraints:
- Format: Use clear markdown headings, numbered steps, and tables.
- Style: Direct, encouraging, a bit daring, but never reckless; emphasize experimentation and reversibility, not all-or-nothing bets.
- Scope: Focus only on 2026; no generic long-term life planning.
- Reasoning: Think through the psychology of fear and reversibility before you generate the plan, but only show the user your final, polished output.

Take a deep breath and work on this problem step-by-step.

What This Uncovers

This prompt uncovers a structured method for turning aspirations into action by identifying bold moves that feel scary but carry minimal permanent risk.

It will reveal to you how to systematically evaluate opportunities and commit to experiments that could transform your life without betting everything on a single irreversible decision.

Prompt 22 —> The Subtraction-Only Quarter

By Mia Kiraki 🎭 - “If I could only grow in Q1 by removing things, what would I dare to delete, merge, or ignore?”

The Prompt

You are a subtraction strategist. Your rule: no new tactics for one quarter—only cuts, consolidations, and focus.

I’ll paste:
1) My 2026 Q1 content + marketing plan (channels, cadence, formats, offers).
2) Any notes on what felt heavy or low-joy in 2025.

Your tasks:
1. Mark each item as:
   - CORE ENGINE: directly tied to revenue or list health.
   - SUPPORTING: nice-to-have, but not critical.
   - NOISE: low differentiation, low return, or pure obligation.
2. Design a “Subtraction-Only Quarter”:
   - At least 3 things to kill or pause.
   - 2 things to merge or simplify.
   - What stays untouched as the core engine.
3. Rewrite my Q1 plan in its minimal, focused form.
4. Write the dare: what I must *not* do in Q1, even when I get anxious.

Output:
- Labeled list (core / supporting / noise)
- New minimal Q1 plan
- A clear “do not touch” / “do not add” dare.

Treat my anxiety as the main obstacle, not my imagination.

What This Uncovers

This forces you to recognize how much of your “strategy” is actually fear-driven clutter that conceals what really moves the business.

Prompt 23 —> The Template Bonfire

By Mia Kiraki 🎭 - “What would my writing look like if I stopped hiding behind my own templates for a few issues?”

The Prompt

You are my format arsonist. Your mission: identify which templates I’ve outgrown and dare me to ride without them.

I’ll paste:
- The current templates or skeletons I use for emails/posts (outline structures, standard sections, recurring openings).
- 3–5 recent pieces that follow them.

Your tasks:
1. Describe the main template patterns I rely on (structure, opening style, CTA placement, etc.).
2. For each pattern, answer:
   - What safety does this give me?
   - What originality or tension does it remove?
3. Choose one upcoming piece and design a “template-free” version:
   - A different structure or narrative shape.
   - A different opening move than I usually use.
4. Write the dare:
   - Which 1–2 issues/posts I must publish in 2026 with NO existing template allowed.

Output:
- My template patterns + their comfort function
- A re-imagined outline for one piece
- A concrete 1–2 issue Template Bonfire dare.

Assume I’m good enough to write without the training wheels—and terrified of proving it.

What This Uncovers

This surfaces how your beloved systems might now be suppressing your range and spontaneity. It nudges you into controlled chaos where fresh structure can unlock a new tier of work.

Prompt 24 —> The Delulu Audit

By AI Meets Girlboss - “Is your content strategy giving main character or background extra? If your brand had to audition for a role in its own content strategy… would it even get a callback?”

The Prompt

Act as a brutally honest casting director. I’ll paste my 2026 content plan below. Audit it like a talentless-but-delusional starlet walked into your studio.
What is giving main-character energy?
What is giving beige?
Where are the plot holes?
What would make this storyline bingeable?
Rewrite my content strategy so it finally has a storyline, a hook, and a lead character worth rooting for.

What This Uncovers

This shows which numbers are quietly steering your behavior toward shallow optimization instead of meaningful growth. It dares you to trade dopamine dashboards for a tiny set of signals aligned with actual outcomes.

Prompt 25 —> The Opinion Ceiling Breaker

By Mia Kiraki 🎭 - “What does my fully unhedged opinion look like—and can I release even a fraction of that into the wild?”

The Prompt

You are my opinion strength trainer. Your goal: raise the ceiling on how directly I’m willing to say what I believe.

I’ll paste:
1) 5–10 “spicy but safe” opinions I’ve published (with links or text).
2) My private, unedited versions of those opinions (journal notes, DMs, rants to friends), if available.

Your tasks:
1. For each opinion pair (public vs. private), describe:
   - What I removed (names, stakes, specificity, blame, self-implication).
2. Choose 2–3 opinions where the gap between private and public is the biggest.
3. For each of those, draft:
   - Version A: 10–15% bolder than my current public version.
   - Version B: 30% bolder, but still fair and evidence-based.
4. Design a 3-post “Ceiling Breaker” sequence:
   - In what order to publish these.
   - How to frame them to my audience so they land as leadership, not meltdown.

Output:
- Gap analysis for each opinion
- 2 graded rewrites for the top 2–3 opinions
- A simple 3-post release plan.

Assume my audience can handle more truth than my ego thinks.

What This Uncovers

The prompt helps you discover if your content has no plot or personality. The model forces a storyline into your strategy, which instantly makes it more memorable and shareable.

EPILOGUE

One year later, the winter ball looks different.​

The robots are still here, but this time there are humans too: builders, creators, makers, all the ones who survived 2025 and didn’t drown in frameworks and AI workflows.

No masks this time. The factory floor is warm and the dome overhead is gone, replaced with open rafters where snow drifts in but doesn’t stick.​

Someone new to the party approaches the punch table where the Foreman is standing. “How’d you all figure it out?” they ask.​

The Foreman pulls a card from its pocket and hands it over. Four lines, handwritten:

1. Always create a context doc;

2. Share a reference file if at all possible;

3. Don’t be afraid to chuck out AI’s first draft;

4. If you’ve gone through 4-5 rounds with AI and haven’t seen any progress, reword your request/rethink your approach.

“Karen Spinner”, the Foreman says. “She gave us this before we got in the sleigh.”

The human reads it twice, then looks around the room. A robot near the fire barrel is throwing away a stack of AI outputs without flinching. Another one at a worktable flips through a thick binder labeled “Context + References” and nods, satisfied.

“It’s the simplest thing,” the silver-masked robot says, walking over. “But we spent all of 2025 pretending we didn’t need it. We thought raw prompts and vibes were enough.” It taps the card. “This is what got us out.”​

Outside, the Robot Reindeers are already harnessed for next year’s run, but they’re not moving yet. There’s no rush. The sleigh will leave when it’s ready, and this time, everyone knows what they’re bringing with them.​

The Foreman looks back through the factory doors one last time, then turns and looks straight at you. “Your turn,” it says.​

You have 25 prompts.

You don’t need to use all of them BUT the one that made your stomach drop. That physical reaction tells you exactly where the growth is hiding.

My challenge to you is simple:

Pick one prompt from whatever Act challenges you most. Share your output in the comments below. OR, if you’re already in a holiday mood, just share your wish for next year.

Let the robots handle the rest. I’ll see you in 2026.

Chief 🤖 at ROBOTS ATE MY HOMEWORK

📅The robots kept the receipts from last year. Here are a few other assignments the machine found worth saving:

Autocomplete is stealing your homework

The secret life of unnamed streets in your AI stack

Build your content like a movie trailer the brain can’t walk away from

An AI that's trained on best practices will never find the dragons