The problem with a new idea is that it arrives fully dressed. You see the finished newsletter, the packed launch, the nice round subscriber number. You do not see the six months of grinding out content nobody reads. The gap between those two pictures is where I’ve lost the most.

So I built a gate. Eight prompts I run before I let myself fall in love. I run them in Gemini, not ChatGPT, and there’s a specific reason for that, which I’ll get to. The short version: Gemini can actually go and check whether the world agrees with me, and I’ve stopped trusting any idea validation that can’t.

Here’s the whole thing.

Why Gemini for this

Most idea-validation prompts are for show. You describe your idea, the model tells you it’s promising, you feel good, you build it, and it dies. The model was never checking anything. It was pattern-matching your enthusiasm back at you.

Gemini’s free tier does two things that break that loop. Its answers are grounded in Google Search, so when I ask whether a market exists, it goes and looks rather than guessing from training data. And Deep Research will spend a few minutes crawling dozens of pages and hand me back a proper report. Free accounts get a handful of those a month, which is plenty when you’re only running it on ideas that survive the first few prompts.

There’s also Gems, which are custom saved assistants, free for everyone now. I’ve turned this whole sequence into one. More on that at the end.

The point of grounding is simple. I don’t want a model that agrees with me. I want one that can be sent to find out I’m wrong.

The eight prompts

Run them in order. Each one is allowed to kill the idea. If it does, you stop, and you’ve saved yourself a season of work.

Read more

I got tired of guessing what to write next. So I handed the problem to ChatGPT, and not in the vague “give me 20 blog ideas” way that produces 20 ideas you’d never publish. I made it audit what I’d already done, find the holes, and fill them with a proper calendar. Three months came out the other side. Most of it was better than what I’d have picked myself.

Here’s the exact sequence.

The idea behind it

The trick isn’t asking for ideas. Anyone can get ideas. The trick is making the model understand your existing mix well enough to spot what’s missing, which is a completely different job.

A gap is only a gap relative to a pattern. So the first half of this is teaching ChatGPT your pattern. The second half is asking it to break the pattern usefully. Skip the first half and you get generic filler. Do it properly and you get ideas that fit your voice but point somewhere you haven’t been.

One note on tiers before we start. The six core prompts run on the free version of ChatGPT. Two optional upgrades near the end use Deep Research and custom GPTs, both of which sit behind the paid plan now. I’ll flag those clearly so you know what’s free and what isn’t.

The six prompts

Run them in one session so the model keeps the context. Each one builds on the last.

1. Feed it the raw material

Don’t describe your newsletter. Show it. Paste in your last 20 to 40 post titles, or the titles plus a one-line summary of each if you have them handy.

Here are my recent newsletter posts, titles and short summaries. Read them as a set, not one by one. Don’t comment yet. Just confirm you’ve got them and tell me you’re ready for the next step.

Making it hold the whole set at once matters. You want it reasoning across the corpus, not reacting to the last thing it read.

2. Make it map what you actually cover

Now group these posts into themes. Give me the real categories my content falls into, not the ones I’d claim. Tell me roughly what share of my output sits in each theme. Be honest if one theme is eating everything.

This step alone is worth the exercise. Seeing your own mix as percentages is uncomfortable in a useful way. I found nearly half my posts clustered in one corner I thought was a sideline.

Read more

Downloadable Claude Skill and Custom GPT

On 21 July, Substack switched on an AI detector. It’s built with a company called Pangram, and it scans posts, notes and comments over 100 words. Anyone can run it. Readers on your work, you on your own drafts, or a stranger on someone they’ve taken against.

It’s opt-in and there’s no penalty attached. Nobody gets ranked down or banned for a high score. Chris Best has been careful to frame it as a first step toward transparency rather than a rule about which tools you’re allowed to use.

I’m relaxed about it, and I should explain why before I get to the interesting part.

I’ve never hidden that I use AI. I wrote an Honest AI Manifesto in April saying exactly how and where. Some pieces here are drafted almost entirely by hand and tidied with a model. Some start as a conversation and get rewritten by me four times. A few are mostly mine with a paragraph I couldn’t crack handed over in frustration. That’s the honest answer and it’s been on the record for months.

So a scanner isn’t a threat to me. It might tell readers something they already know.

I do have one worry, and it isn’t about my score. The false positives land in a pattern. They hit people writing in a second language, and they hit neurodiverse writers whose rhythm and repetition a model reads as synthetic. One writer ran the same piece twice last week and got 100% AI, then 100% human. That’s a coin flipper, not a detector.

But that’s an argument for treating the number with suspicion, not for hiding from it.

Slop is a readability problem before it’s a moral one

Using AI is fine. Publishing what it hands you first time is where it goes wrong.

Generated prose has a texture. Every sentence lands between 15 and 20 words. Every section closes with a tidy summary. The same eighty verbs turn up over and over. Delve. Leverage. Foster. Navigate. Nothing is technically wrong, and that’s the trouble, because there’s nothing to catch on.

Human writing is lumpy. Two-word sentences next to forty-word ones. A paragraph that goes on too long because the writer got interested. A joke that doesn’t quite land. Repetition where they should have varied it. Those are the warts, and readers need them. They’re the handholds.

Strip them out and you get text that’s correct and frictionless, which sounds like a compliment until you notice nobody finished reading it.

So the case for humanising your drafts has nothing to do with detectors. Slop is boring, and boring loses subscribers faster than any label ever will.

What I built

A tool that scans a draft, tells you what’s wrong with it, then edits. You paste your writing in. That’s the whole interface.

The scan counts rather than guesses. Ask any AI to tell you if your writing sounds robotic and it’ll have an opinion, and the opinion will be different tomorrow. This one comes back with numbers, and they’re the same numbers every time.

It catches what your eye slides past. Em dashes. Semicolons. Roughly 300 words and phrases that show up constantly in generated text and hardly ever in good writing. It gets the structural habits too, like the self-posed question (”The result? Devastating.”) and the trailing clause that adds nothing (”...solidifying its position as a market leader”).

Then it measures rhythm. It works out how much your sentence lengths vary, and your paragraph lengths, and tells you when the answer is “not enough”.

You get a number at the end. Under 10 per thousand words reads human. Above 30 and the piece needs rebuilding rather than tidying.

I ran a deliberately awful sample through it and got 48. I ran a passage I was proud of and got zero.

What it won’t do

It won’t help you beat Pangram, and I’ve built it to refuse if you ask.

The refusal is practical. The job can’t be done reliably. Detectors score statistical patterns in how text was generated, which isn’t the same thing as the vocabulary and sentence shapes this edits. The two overlap a bit. Your score might drop after an edit. It might go up. Anyone selling you a guaranteed number is selling you a guess in a confident voice.

And chasing the score is the wrong game anyway. If you’re editing to satisfy a classifier, you’ll produce writing that satisfies a classifier. Nobody has ever subscribed to a newsletter because it scored well on anything.

Edit so it reads better. The score does what it does.

The part many ‘humanising’ tools get wrong

Every de-slop tool I’ve tried is too keen.

You write about actual landscapes and it strips out “landscape”. You use “systemic” correctly in a piece about systemic risk and it swaps in something vaguer. You make a joke and it sands the joke off, because jokes are irregular and irregular looks like a mistake.

What comes back is clean, correct and dead. Which is the same problem you started with, wearing a different coat.

So this one shows its work. Every edit comes with a change log, and underneath that, a list of everything it flagged and deliberately kept, with reasons. If it left your weird metaphor alone, it says so and says why. You can disagree with it, which is more than most editing tools allow.

Where to get it

It’s a Claude Skill, so it loads by itself when you mention humanising a draft. There’s a ChatGPT version built as a custom GPT. It runs the same scanner under Code Interpreter. The Gemini version works but loses the diagnostic, because Gems can’t run code. I’d rather say that now than have you find out.

The term list is a plain text file. Add your own words, delete mine, and the scanner picks up the change without anyone touching code. My list bans “delve” and “tapestry” and about 298 other things. Yours will be different. Mine bans “utilise” purely out of spite.

Download the files for Claude and ChatGPT, including instructions, here (zip file - unzip when downloaded) …

Read more

It’s inbox summarisation. Someone gets AI access, opens their email, and thinks: this is the obvious place to start. Forty unread emails become a tidy bullet list in ten seconds. It feels like a win.

It isn’t. Summarising your inbox doesn’t remove any of the work. You still have to read the original emails to reply properly, you still have to decide what matters, and now you’ve added a summary you have to check against the source. The actual bottleneck in most people’s day isn’t reading email. It’s deciding what to do about it and writing the reply. That’s the part most workflows skip, because it’s harder and less satisfying to set up.

This article gives you 8 prompts that target the real bottleneck instead. You’ll go from “summarise this” to “draft my actual response, flag what needs a decision, and tell me what I can ignore.”

Why this matters now

Most AI workflow advice still treats summarisation as the entry point, because it’s the easiest demo. It looks impressive in five seconds and it’s the first thing every tool tutorial shows you. But a workflow only saves time if it removes a step. Summarising adds one. The prompts below skip straight to decisions and drafts, the parts of your day that actually eat the hours.

Prompt 1: The reply drafter

What it does: Writes a complete first-draft reply to an email, ready for you to edit and send.

When to use it: For any email that needs more than a one-line response. Skip it for anything you can answer in five words.

The Prompt:

Here’s an email I need to reply to. Write a complete draft reply in my voice: direct, no corporate filler, short paragraphs. Address every question or request in the original email. If something needs more information from me before you can answer it properly, flag that clearly at the top instead of guessing. Email: [PASTE EMAIL]

How to use it:

1. Open the email you’d normally summarise and move past

2. Paste the full text into the prompt

3. Edit the draft for tone and send, rather than starting from a blank reply box

Example input: Here’s an email I need to reply to. [pasted client email asking for a project update and a new deadline]

What you’ll get: A complete reply addressing both the update and the deadline question, with a flag if you haven’t given it enough detail to answer the deadline part confidently.

Advanced note: Save 3 or 4 of your own past replies somewhere and paste one in alongside the prompt occasionally. It keeps the draft sounding like you instead of drifting into generic email voice over time.

That one prompt alone removes the blank-page problem on every reply you write this week.

But replying faster is only half the fix. The other half is knowing which emails need a decision from you and which ones don’t need anything at all. The next 7 prompts handle that.

Here’s what’s behind the paywall:

Prompt 2, the decision flagger: Reads a batch of emails and tells you which ones actually need a choice from you, not just a reply.

Prompt 3, the no-reply filter: Identifies emails you can archive or ignore without any response, and tells you why.

Prompt 4, the meeting request handler: Drafts accept, decline, or reschedule replies to meeting requests based on your stated priorities.

Prompt 5, the follow-up tracker: Scans a thread and tells you what you’re still waiting on from someone else, so nothing falls through.

Prompt 6, the tone matcher: Adjusts a drafted reply to match the formality level of the person you’re writing to.

Prompt 7, the batch responder: Drafts short replies to several low-stakes emails at once, so you clear the small stuff in one pass.

Prompt 8, the weekly inbox review: Looks back at a week of email and tells you which relationships or threads need attention you’ve been missing.

Plus: a one-page workflow audit framework, a short set of questions to run against any AI workflow before you build it, so you catch this mistake yourself next time instead of reading about it after the fact.

If you only do one thing this week, run Prompt 1 on your next three emails. If you want the system that actually clears your inbox instead of just describing it to you, the rest is one upgrade away.

Upgrade to get all 8 prompts, the workflow audit framework, and the full step-by-step sequence

Read more

There’s a difference. Comments are generally where people agree. Complaints are where they tell you what’s actually broken in their work or their process. I think the majority of people scrolling LinkedIn skim the comments and skip the complaints, because complaints are usually buried three replies deep in a thread about something else entirely.

And that’s the gap. Everyone in your niche is publicly narrating their problems on LinkedIn right now, in posts, in comments, in the replies to other people’s posts. No amount of guessing or formal market research delivers this intel as fast or as cheaply.

This article gives you 8 prompts that turn LinkedIn into a research tool. You’ll pull real language from real posts, sort it into patterns, and end up with a shortlist of problems worth solving, and crucially, in words your audience already uses.

Why this works now

Like all social platforms, LinkedIn’s algorithm rewards posts that get engagement. This means the platform is full of people venting, asking, and comparing notes in public. That’s free research sitting in plain sight. AI makes it usable because it can read fifty posts and group the patterns in minutes, something that would take you an entire afternoon by hand. You don’t need a survey. You need fifteen minutes and the right prompt.

Prompt 1: The struggle scrape

What it does: Pulls the recurring problems mentioned across a batch of LinkedIn posts or comments from your niche.

When to use it: Right at the start, before you do anything else. This is your raw material.

The Prompt:

I’m going to paste a batch of LinkedIn posts and comments from people in [YOUR NICHE]. Read through all of them and pull out every specific problem, frustration, or struggle mentioned, even small ones. For each one, quote the exact phrase they used and note who said it (just their role or title, not their name). Don’t summarise or interpret yet. Just extract. Posts: [PASTE POSTS]

How to use it:

1. Go to LinkedIn and search a keyword or hashtag tied to your niche

2. Copy 15 to 20 posts and their top comments into a doc

3. Paste the lot into the prompt and run it

Example input: I’m going to paste a batch of LinkedIn posts and comments from people in freelance copywriting. [20 posts pasted]

What you’ll get: A list of 30 to 50 quoted frustrations, each tagged with the poster’s role. Some will repeat. That’s the point.

Advanced note: Don’t clean the posts before pasting them. Typos, ALL CAPS, and rambling are signal. AI will smooth them out if you ask it to summarise too early, and you’ll lose the texture that makes a problem feel real.

That prompt alone will hand you a raw list of problems straight from the people who have them, no guessing required.

But one list of quotes is just that, a list, not a strategy. The next 7 prompts turn that list into something you can actually act on.

Here’s what’s behind the paywall:

Prompt 2, the pattern sorter: Groups your raw quotes into 5 to 8 actual themes instead of 40 scattered complaints.

Prompt 3, the frequency ranker: Tells you which problems show up again and again versus which ones are one-off noise.

Prompt 4, the language miner: Extracts the exact words and phrases your audience uses, so your content stops sounding like a textbook.

Prompt 5, the hidden problem finder: Surfaces the struggles people imply but never say outright, the ones competitors miss.

Prompt 6, the content angle generator: Turns each pattern into 3 ready-to-use post or email angles.

Prompt 7, the audience segment splitter: Shows you which struggles belong to which sub-group inside your niche, so you stop writing to an imaginary “everyone.”

Prompt 8, the validation check: Cross-checks a problem you think you’ve found against a fresh batch of posts, so you’re not building on a fluke.

Plus: a one-page pain language tracker, a simple log for the exact phrases people use, so the patterns build across weeks instead of resetting every time you do this.

If you only do one thing this week, run Prompt 1 on twenty posts and read what comes back. If you want the system that turns that into a repeatable research habit, the rest is one upgrade away.

Upgrade to get all 8 prompts, the pain language tracker, and the full step-by-step sequence.

Read more

Which best describes your working week?

a) Publishing. Newsletter, blog, videos, building an audience.
b) Marketing, for clients or an employer.
c) Running a small business. Publishing is one job among several.
d) A job inside a larger organisation.

Just hit reply and send a, b, c or d. Nothing else needed.

Cheers
Andy

They’re getting better at writing prompts. Tighter instructions, cleverer personas, more specific examples. And the outputs do get better, for a while, on the tasks they thought to test. The problem is that “it looked good when I tried it” is not a quality standard. It’s a hope.

Tools like DSPy are already optimising prompt text automatically. The writing part is being automated. What can’t be automated, at least not yet, is knowing whether the output is actually good. That requires a human to define what good looks like, write down the test cases, and run them consistently.

That’s the skill shift. From writing prompts to testing them.

These 9 prompts take you through the full process: building test cases, scoring outputs, finding failure modes, running comparison tests and locking in the version that actually works.

Why this matters right now

Automated prompt optimisation is moving fast. DSPy, PromptFoo, and similar tools can iterate through hundreds of prompt variants in the time it takes you to write one manually. The bottleneck they all share is the same: someone has to tell the system what a good output looks like.

If you can’t answer that question in writing, with specific criteria, the tools are useless. The people who’ll get the most from the next wave of AI tools aren’t the best prompt writers. They’re the best prompt evaluators.

Prompt 1: Build your first test case

What it does: Produces a structured test case document for any prompt you’re currently using, including input variations, success criteria and failure flags.

When to use it: Before you rely on any prompt for real work, or when an existing prompt is producing inconsistent results.

The Prompt:

I want to build a proper test case for an AI prompt I’m using. Here is the prompt I want to test:

[PASTE YOUR PROMPT HERE]

My use case is: [DESCRIBE WHAT YOU USE THIS PROMPT FOR, e.g. “writing LinkedIn posts for a B2B software company”]

My target output quality is: [DESCRIBE WHAT GOOD LOOKS LIKE, e.g. “under 200 words, no jargon, ends with a question, sounds like a person not a brand”]

Please produce a test case document with: 1. 5 input variations that cover the range of real scenarios I’d use this prompt for (from easy/typical to difficult/edge case) 2. For each input, a written description of what a passing output would look like 3. A list of 5 specific failure flags (things an output could do that would make it a fail regardless of other quality) 4. A scoring rubric with 4 criteria, each scored 1-5, that I can apply to any output from this prompt

Format the result so I can paste it into a document and reuse it.

How to use it:

1. Paste your existing prompt into the first placeholder exactly as you currently use it

2. Write the use case description in plain language, not aspirationally, describe what you actually use it for

3. For target output quality, be specific: length, tone, structure, what it must and must not contain

Example input: Prompt = a prompt that writes cold email subject lines. Use case = “outreach emails for a freelance UX designer targeting e-commerce startups.” Target quality = “under 8 words, no question marks, specific to the recipient’s industry, doesn’t use the word ‘quick.’”

What you’ll get: A reusable test document with 5 scenario inputs, pass/fail descriptions for each, a list of automatic failure conditions and a 4-criteria scoring rubric. Takes about 3 minutes to fill in and saves hours of guesswork later.

Advanced note: Run the same test case document through two different models (Claude and GPT-4o, for instance) using identical inputs. The differences in where each model fails tell you more about prompt weaknesses than any amount of manual tweaking will.

ooops, a paywall…

That test case gives you something most people never have: a written definition of what good actually means for your specific prompt.

Behind the upgrade, there are 8 more prompts that take this further:

Prompt 2: Score an existing output: Apply your scoring rubric to a real output and get a structured evaluation with a numerical score and specific improvement notes, not just “this could be better.”

Prompt 3: Find failure modes: Run your prompt against 6 adversarial inputs designed to break it, so you know where it falls apart before your actual work does.

Prompt 4: Write a comparison test: Set up an A/B test between two prompt variants using identical inputs and a blind scoring method, so you can make version decisions on evidence rather than preference.

Prompt 5: Extract implicit criteria: Feed Claude a set of outputs you rated highly and ask it to reverse-engineer your unstated preferences into explicit criteria you can add to your rubric.

Prompt 6: Build a regression log: Create a living document that records prompt versions, test scores and change notes, so you can roll back to a previous version if a tweak makes things worse.

Prompt 7: Stress-test for edge cases: Generate 10 edge-case inputs specific to your use case, including unusual formats, missing information and off-topic requests, to find where your prompt gets confused.

Prompt 8: Write a minimum viable spec: Turn your test case document into a one-page prompt specification that another person (or an automated tool like DSPy) could use to optimise or rebuild your prompt from scratch.

Prompt 9: Audit a prompt you didn’t write: Apply the full testing process to a prompt you found online or inherited from someone else, so you can decide whether it’s actually fit for purpose before using it in production.

Plus: The Eval Scorecard, a reusable plain-text template you can copy into any document and fill in for any prompt in under 10 minutes.

If you only do one thing differently this week, building that first test case is a solid start. The full testing workflow is behind the upgrade.

Read more

I knew there had to be a better way. Turns out there was, and it had been sitting in Claude’s settings the whole time.

Skills files. A persistent instruction set that loads automatically whenever you open a chat. You write it once, Claude knows your defaults forever. No preamble, no re-explaining, no “as I mentioned earlier.”

I spent a couple of hours building one for copywriting work. The prompts below are exactly what I used. The result is a version of Claude that already knows your clients, your process and your rates before you type a single word.

Why Skills files aren’t just for developers

Most of what’s written about Skills online is aimed at coders. Which makes sense. They got there first.

But a developer’s work is fairly self-contained per session. Copywriting isn’t. It runs on accumulated context: who the client is, how they sound, what a good brief looks like, what you’re going to push back on. That context disappears every time you close the tab.

Skills is where you put it so it doesn’t.

Prompt 1: Brand voice entry builder

What it does: Produces a fully formatted Skills entry for a specific client’s brand voice, ready to paste straight into your Skills file.

When to use it: When onboarding a new client, or when you’ve been working with someone long enough that you can describe how they sound with some precision.

The Prompt:

You are helping me build a Claude Skills file entry for a client’s brand voice. I’ll give you information about this client and you’ll turn it into a clean, structured Skills entry I can paste directly into my file.

Client name: [CLIENT NAME] Industry: [INDUSTRY] Target audience: [WHO THEY’RE WRITING FOR] Tone description: [HOW THEY SOUND - e.g. “direct but warm, plain English, no jargon”] Words or phrases they use: [SPECIFIC VOCABULARY OR PHRASES] Words or phrases to avoid: [THINGS THAT DON’T FIT THEIR VOICE]Output types I write for them: [E.G. “email sequences, landing pages, LinkedIn posts”] One example of their writing I like: [PASTE A SHORT SAMPLE]

Format the output as a Skills entry with these sections: Voice overview (2–3 sentences), Tone markers (bullet list), Vocabulary rules (bullet list), Output defaults (what to assume unless told otherwise). Keep it under 300 words.

How to use it:

1. Fill in each placeholder from memory, not from a mood board. If you can’t describe how the client sounds without looking something up, you don’t know their voice well enough yet.

2. Read the output like a client would. Does it sound like them? Edit anything that’s off before you file it.

3. Paste it into your Skills file under a heading like “Client: [Name] - Brand voice.” Keep that heading consistent so Claude can find it.

Example input: Client name: Hartley & Co / Industry: B2B SaaS / Target audience: ops managers at mid-size logistics firms / Tone: plain-spoken, slightly dry, no marketing fluff / Words they use: “straightforward,” “works out of the box,” “no surprises” / Words to avoid: “innovative,” “cutting-edge,” anything with an exclamation mark / Output types: case studies, email sequences, web copy / Example: “Hartley tracks your fleet in real time. No setup headaches, no ongoing IT support. It just works.”

What you’ll get: A 200-250 word Skills entry with four clearly labelled sections, formatted for direct paste into your Skills file.

Advanced note: If you write for the same client across different channels (e.g. long-form articles versus social posts), run this prompt twice with different output types specified. The voice entry will pick up the register shifts between channels and give you a more accurate default.

oops, you hit the paywall!

Here’s what’s behind it…

The bonus is a full Skills file template with all seven sections pre-labelled and placeholder text in every slot. You can open it, work through the prompts in order, and have a complete file by the end of the session.

Plus the other seven prompts in the sequence:

Prompt 2 - Brief intake entry builder: Produces a Skills entry that tells Claude exactly what information you need from a client before you start writing, so it can flag gaps automatically when you paste a brief in.

Prompt 3 - Platform tone calibrator: Builds a Skills entry for output defaults per platform: what changes between a LinkedIn post, an email subject line and a homepage hero when you’re writing for the same client.

Prompt 4 - Feedback translator: Creates a Skills entry that helps Claude turn vague client feedback (”make it punchier,” “less salesy”) into specific edits rather than a full rewrite.

Prompt 5 - Scope and rate defender: Builds an entry that gives Claude enough context about your pricing to help you write scope-of-work language and push back on feature creep.

Prompt 6 - Portfolio case study writer: Produces a Skills entry with your preferred case study format, so Claude can draft a first version from a few bullet points about a completed project.

Prompt 7 - Client onboarding doc generator: Builds an entry covering your working process, turnaround times and revision policy, so Claude can produce a ready-to-send onboarding doc for any new client.

Prompt 8 - Generic output checker: Creates a Skills entry with your personal quality bar, so you can run any draft through Claude and get a specific list of what needs fixing and why.

Read more

The problem usually isn’t the prompt. It’s what you asked for.

Power users have worked out, mostly through throwing away a lot of drafts, that certain requests almost guarantee that assembled feeling. The ask itself is the problem. You’re pointing Claude at the part of its training where everything sounds the same.

Six of those requests. All common. All worth stopping.

1. “Write me an introduction”

The introduction is the part Claude is worst at, and it’s the first thing most people ask for.

Ask Claude to write an introduction and you’ll almost always get throat-clearing: a sentence that restates the topic, a sentence that explains why it matters, and a transition into the main content. It reads like a five-paragraph essay format got applied to whatever you’re writing.

What power users do instead: they write the introduction themselves, or they ask Claude to write the second paragraph first.

The reason this works is that Claude’s defaults for openings are pulled from an enormous body of content where openings are formulaic. The middle of a piece has more variation. Start there, and the opening you write to connect to it will feel more natural because you’ve already got something real to connect to.

If you want Claude to write an opener, give it the whole piece first and ask: “Write a one or two sentence opening that drops straight into the argument. No context-setting, no explaining what the piece is about. Assume the reader already knows why they’re here.”

That constraint breaks the formula.

That fix alone changes what you get back. The constraint at the end of that prompt (”no context-setting, no explaining what the piece is about”) is doing more work than most full rewrites would.

The next five mistakes are less obvious, and a couple of them are genuinely counterintuitive. Subscribers get:

Mistake 2 — “Make it more engaging”: Why vague quality instructions produce vague quality improvements, and the specific questions that actually fix flat writing.

Mistake 3 — “Give me 10 ideas”: The padding problem that gets worse the higher the number, and the single-question move that produces better output than any list.

Mistake 4 — “Write me a summary”: Why summary requests produce useless output, and how naming the use case and audience changes everything you get back.

Mistake 5 — “Does this sound okay?”: The feedback request that almost always produces reassurance with a garnish, and how to ask for actual criticism instead.

Mistake 6 — “Write it in my voice”: Why naming your voice gives Claude nothing to work from, and the showing-not-naming approach that produces output that sounds like a specific person wrote it.

Plus: the constraint swipe file, six phrases that replace vague instructions with something Claude can act on.

Read more

Eight setups. Each one takes under 20 minutes to build. After that, the heavy lifting on briefs, copy quality, and reporting mostly runs in under ten minutes per task.

Skill 1: The brand voice analyst

What it does: Gives you a reusable voice profile for any client, built from their existing content.

When to use it: First week of any new engagement, before you write a single word on their behalf.

You are a brand voice analyst. I’m going to give you a sample of content from a client. Analyse it and produce a voice profile with these sections: tone (3-4 adjectives with short explanations), sentence structure patterns, vocabulary preferences and words to avoid, things they never say, and a short “write like this” instruction set I can paste into future prompts. Here’s the content: [PASTE 500-1000 WORDS OF CLIENT CONTENT]

How to use it:

1. Paste in a mix of their best-performing content (emails, web copy, LinkedIn posts).

2. Save the output as a “voice file” in a Claude Project for that client.

3. Reference it at the top of every future prompt: “Use the voice profile below.”

Example input: Three LinkedIn posts + one email newsletter from a B2B SaaS client.

What you’ll get: A repeatable voice spec you can use across every deliverable. Two consultants I know have started including this as a paid discovery item in their onboarding.

Advanced note: Run the same prompt on a competitor’s content. The gap between how your client talks and how their competitors talk often surfaces a positioning angle worth flagging in strategy work.

That setup alone is worth building in week one of any engagement. A voice profile you can paste into every future prompt changes the baseline quality of everything that follows.

The next seven skills are where the billable leverage actually lives. Subscribers get:

Skill 2 — The messaging stress-tester: Puts your client’s draft copy through a sceptical B2B buyer read before it goes anywhere near a decision-maker, with specific weak spots identified rather than a vague “this doesn’t feel right.”

Skill 3 — The content repurposing engine: Turns one piece of long-form client content into six ready-to-publish formats in a single run, using the voice profile from Skill 1.

Skill 4 — The competitive intelligence brief: Produces a structured competitor snapshot in about eight minutes per rival, including one gap they’ve left open and one thing your client should stop doing because the competitor does it better.

Skill 5 — The client onboarding questionnaire builder: Writes a tailored discovery questionnaire from any brief, grouped and annotated, ready to send 48 hours before kick-off.

Skill 6 — The campaign angle generator: Produces six genuinely distinct campaign angles from a single brief, each with a built-in weakness flag that tends to save a lot of meeting time.

Skill 7 — The crisis comms first-responder: Generates a holding statement, internal communication, and a “do not do” list within minutes of a reputational issue landing, before you call the client back.

Skill 8 — The board-ready report formatter: Turns raw working notes into an executive summary structured for a non-marketing CEO, clean enough to forward directly.

Plus: the consultant’s Claude swipe file, six prompt openers that set up the right specialist context before you add the specific task.

Read more

Most productivity advice says to get better at prioritising. That’s true but useless without a system that makes prioritisation fast. These seven prompts give you one. You dump the chaos in. You get a ranked, scoped, executable plan out. The whole sequence takes under 20 minutes.

Prompt 1: The brain dump processor

What it does: Takes a raw, unfiltered dump of everything in your head and sorts it into four clean buckets: projects, tasks, ideas worth exploring, and things you can delete entirely.

When to use it: First. Before any other prompt in this sequence. You need to get it out of your head and into a format you can work with.

I’m going to give you a raw brain dump. It will be messy, incomplete, and probably contradictory. Your job is to read everything I’ve written and sort it into four categories: (1) Projects, things that require more than one action to complete, (2) Tasks, single actions I could do in one sitting, (3) Ideas worth exploring, things that might become projects but aren’t ready yet, (4) Delete candidates, things that have been on my list for more than a month and I haven’t acted on. For each item, put it in the most fitting category and add a one-line note on why. Don’t ask me clarifying questions. Make a call on everything. Here’s the dump: [PASTE YOUR FULL BRAIN DUMP]

How to use it:

1. Open a blank document and write everything down without filtering. Set a timer for ten minutes if that helps. Include half-formed thoughts, things you’ve been meaning to do for six months, and anything else taking up mental space.

2. Paste the whole thing. Don’t tidy it up first. The messier the input, the more useful the sort.

3. Look at the “delete candidates” list first. Anything you feel relieved to see there is gone.

Example input: “Rewrite website, finish the course outline, email Sarah about the collab thing, that podcast idea I had in March, update my LinkedIn, launch the community, write the case study I promised Jake, the SaaS tool idea, fix the proposal template, read that book on pricing, start a YouTube channel, the newsletter rebrand, do my tax stuff, the coaching programme idea, finish reading Stripe docs...”

What you’ll get: Four clean lists with a brief rationale for each item’s placement. Most people find the “delete candidates” list alone is worth the ten minutes. Getting twenty things off your mental register in one pass changes how the rest of the list feels.

Advanced note: If an item appears in both “projects” and “ideas worth exploring” in your mind, the prompt will force a decision. If you disagree with the call it makes, move the item manually. The disagreement itself is useful information about where your actual priorities are.

That prompt gets everything out of your head and into four categories you can actually act on. Most people find the delete candidates list alone is worth the ten minutes.

The next six prompts turn that sorted list into a working plan.

Prompt 2 — The priority ranker: Scores your projects list on impact, effort, and time sensitivity so you’re picking what to work on next by actual criteria, not by what you feel like doing.

Prompt 3 — The project scoper: Breaks your top-ranked project into a phased action plan with time estimates, dependencies, and one specific thing you can do in the next 30 minutes.

Prompt 4 — The stalled project restart: Diagnoses why something you started has gone quiet, ranks the three most likely reasons, and produces a restart plan from the most probable one.

Prompt 5 — The idea parking system: Turns the ideas that aren’t ready to become projects into a structured holding file, with conditions for when each one is worth revisiting and a trigger date to check them.

Prompt 6 — The weekly reset: Builds a day-by-day plan that protects time for your main project first and tells you clearly what doesn’t fit, rather than pretending everything will.

Prompt 7 — The chaos prevention system: Designs a weekly capture and processing routine around how you actually work, not an idealised version of it, so the brain dump gets shorter every time you run it.

Plus: the single-question triage for when a new idea arrives and you don’t have time for the full sequence.

Read more

“This is for a job interview I really need.” “My manager is presenting this tomorrow and I can’t let them down.” “I’ve been working on this business for three years. Please help me get this right.”

It sounds absurd. The model doesn’t have feelings. It doesn’t care about your career. And yet people keep reporting that it works, and when I test it myself, it often does.

This article doesn’t pretend that’s not strange. What it does is try to explain why it happens, when it actually helps, when it’s just noise, and how to use it without sliding into something that feels genuinely uncomfortable.

Why it works at all

The short version: language models are trained on human text, and human text contains a consistent pattern where high-stakes requests get more thorough, careful responses. When you signal stakes, you’re not appealing to the model’s emotions. You’re activating a pattern in the training data.

A doctor writing to a colleague about an urgent case writes differently than someone dashing off a casual note. A lawyer drafting a contract for a major deal writes differently than someone writing a quick summary. The model has seen millions of examples of this. When you tell it the stakes are high, it shifts register toward the high-care version of whatever you’re asking for.

That’s the mechanism. It’s not magic and it’s not manipulation in any meaningful sense. It’s pattern activation.

Where it gets murky is when you start manufacturing stakes that don’t exist, or when you lean on it so heavily that you stop thinking about what you’re actually asking for.

Prompt 1: The stakes declaration

What it does: Adds a genuine high-stakes frame to your prompt that shifts the model toward a more careful, thorough output.

When to use it: When the default output feels generic or undercooked and you need the model to treat the request as something that actually matters.

[YOUR NORMAL PROMPT]. Before you respond, I want to give you some context: [GENUINE STAKES STATEMENT, E.G. “I’m presenting this to my company’s board next week and the decision will affect whether we get budget to continue” OR “This is the first piece of writing I’m putting my name on publicly and I want it to be right”]. With that in mind, please give me your most careful, considered response rather than the first version that comes to mind.

How to use it:

1. Write your normal prompt first, exactly as you would without this technique.

2. Add the stakes statement after. Keep it to one or two sentences and make it true.

3. The final line (”most careful, considered response”) matters. It’s an explicit instruction, not just emotional colour.

Example input: Normal prompt: Write an about page for my consulting business. I help early-stage startups with go-to-market strategy. Stakes: I’m relaunching my website after two years of freelancing and this page is the first thing potential clients will read.

What you’ll get: A noticeably more considered draft. The model tends to avoid the generic “passionate professional with years of experience” template and pays more attention to specifics when it understands something real is riding on the output.

Advanced note: The stakes have to be real. Not because the model can verify them, but because vague or obviously manufactured stakes produce vague prompts. “This is really important to me” tells the model almost nothing. “My co-founder and I are pitching to three investors on Thursday and this is our exec summary” gives it actual context to work with.

That prompt works. What’s less obvious is when it’s doing real work and when you’re just manufacturing pressure that doesn’t mean anything.

The next six prompts get into that.

Prompt 2 — The effort signal: Tells the model what you’ve already built and invested, which tends to produce outputs that work within your thinking rather than replace it with something generic.

Prompt 3 — The reputation frame: Activates a more specific, considered register by telling the model the output will go out under your name to a real audience you’ve described.

Prompt 4 — The direct callout: Names the failure in a previous response specifically, so the second attempt is a genuine reset rather than a minor variation on the same underperforming output.

Prompt 5 — The manufactured stakes check: This one runs on you, not the model. It’s a self-check for whether the emotional framing you’re about to use is legitimate context or something closer to a lever you’d rather not examine too closely.

Prompt 6 — The precision escalation: Gets the same output quality shift that emotional framing produces, but arrives there through specific context rather than pressure. The cleaner version of the whole technique.

Prompt 7 — The honest brief: Combines stakes, effort signal, and specific context into a single opening. The one worth building the habit around.

Plus: the four-question context checklist that produces better output than any emotional

Read more

The problem is volume. A channel with any traction has thousands of comments and no obvious way to turn them into a content plan. You end up skimming, pattern-matching by gut feel, and making the same type of content you’ve always made.

These seven prompts change that. Copy a batch of comments, paste them in, and you’ll have a prioritised content angle list inside ten minutes.

Prompt 1: The raw comment mine

What it does: Scans a raw dump of YouTube comments and extracts every distinct content signal: questions asked, frustrations expressed, topics mentioned, and things people say they wish the video had covered.

When to use it: Before anything else. This is the starting point for the whole sequence.

You are a content strategist analysing YouTube comments to find content opportunities. Below is a batch of comments from a YouTube video about [VIDEO TOPIC]. Read every comment carefully. Extract and group the following: (1) questions viewers asked that weren’t answered in the video, (2) frustrations or problems mentioned, (3) topics or subtopics viewers brought up themselves, (4) things people said they wanted to see next or wished had been included. Present each category as a plain list. Do not summarise or editorially comment on the patterns yet. Just extract. [PASTE COMMENTS HERE]

How to use it:

1. Go to any YouTube video relevant to your niche. Doesn’t have to be your own channel.

2. Copy 50 to 100 comments. Sort by “Top comments” for the highest-signal batch.

3. Paste them in place of [PASTE COMMENTS HERE].

4. Fill [VIDEO TOPIC] with a plain description, e.g. “AI writing tools for content creators.”

Example input: Video topic: AI writing tools for content creators. Comments: [100 comments pasted from a popular video on the topic]

What you’ll get: Four clean lists that turn an unstructured comment section into readable signal. This is your raw material for every prompt that follows.

Advanced note: 100 comments is a good working size. Under 30 and the patterns are too thin. Over 200 and you start getting repetition that dilutes the output. If the video has thousands of comments, run the prompt twice with different batches and compare.

That prompt turns a wall of unstructured comments into four clean lists you can actually work from.

The next six prompts take those signals through to a finished content plan.

Prompt 2 — The pattern spotter: Groups the extracted signals by theme and ranks them by frequency, so you can see which questions and frustrations kept coming up rather than guessing at it.

Prompt 3 — The content angle generator: Turns the top recurring themes into specific publishable angles, each with a working title written the way a reader would search for it, not the way a content marketer would pitch it.

Prompt 4 — The gap finder: Compares your chosen angle against what already exists on YouTube for that topic, and identifies the specific thing that’s missing before you commit to making it.

Prompt 5 — The hook builder: Uses the actual language, phrases, and complaints from the comments to write opening lines that feel like you’ve had the exact same experience as your reader.

Prompt 6 — The title test: Generates eight title options drawn from real comment language, picks the three strongest, and explains in one sentence why each works.

Prompt 7 — The multi-video synthesis: Runs the analysis across multiple videos and surfaces only the themes that appear consistently across different comment sections, which is where your safest content bets are.

Plus: the comment collection shortcut for pulling batches without copying manually.

Read more

Six months later, you’re resentful, underpaid, and still not sure how to handle the next version of the same conversation.

This isn’t a confidence problem. It’s a language problem. Most freelancers and founders never learned how to decline, push back, or redirect without sounding cold or defensive, because there’s no script for it. These 10 prompts give you one.

Prompt 1: The scope creep redirect

What it does: Writes a response that acknowledges a client’s new request, names the scope boundary clearly, and opens a conversation about next steps, without any trace of annoyance.

When to use it: A client emails asking for something that falls outside the original agreement, and you need to respond before the expectation hardens.

You are a communications writer helping a freelancer respond to a client. The freelancer has a project agreement that covers [ORIGINAL SCOPE]. The client has just asked for [NEW REQUEST], which falls outside that scope. Write a short, warm email response that: (1) acknowledges the request positively, (2) names the scope boundary without using the word “no,” (3) offers two options, either adding it as a paid addition or deferring it to a future phase, and (4) ends with a question that moves the conversation forward. Tone: direct, professional, and zero guilt. No apologising for the boundary.

How to use it:

1. Fill in [ORIGINAL SCOPE] with a plain description of what you agreed to, e.g. “a five-page website with copywriting.”

2. Fill in [NEW REQUEST] with exactly what the client asked for.

3. Copy the output, read it out loud, and cut anything that sounds defensive.

Example input: Original scope: Monthly social media management, 12 posts per month across Instagram and LinkedIn. New request: “Can you also handle our email newsletter going forward?”

What you’ll get: A two or three-paragraph email that redirects without refusing. The client feels heard. You’ve named the line without making it a confrontation.

Advanced note: If the relationship is ongoing and this is the third or fourth scope creep request, add this to the prompt: “The client has made similar out-of-scope requests before. The tone should still be warm but slightly firmer, with a clear implication that any new work needs a new agreement first.”

That prompt handles the request that’s already landed in your inbox. The next nine cover what comes after, and the harder conversations you’ve been putting off.

Prompt 2 — The price pushback response: Holds your quoted rate without apologising for it, with one sentence of value restatement and a clear offer or exit.

Prompt 3 — The “not the right fit” decline: Declines a project inquiry cleanly, without the vague “unfortunately at this time” hedge that leaves people confused about whether you mean it.

Prompt 4 — The retainer renegotiation: Proposes a rate increase or scope reduction on an existing arrangement, framed as a normal business conversation rather than a demand.

Prompt 5 — The “quick favour” deflection: Responds to a request for unpaid advice or a “five-minute call” you know won’t be five minutes, with a redirect or a clean no.

Prompt 6 — The deadline extension pushback: Handles a client trying to compress your timeline, with a clear statement of what’s actually possible and what it would cost to rush.

Prompt 7 — The mid-project “no” to a pivot: Addresses a client who wants to change direction partway through, naming what’s already been done and presenting two options before any more work starts.

Prompt 8 — The long-term client boundary reset: Resets working habits that have drifted over time, framed as practical preferences rather than a list of grievances.

Prompt 9 — The “yes I said yes but actually no” reversal: Walks back a commitment made too quickly, with a specific alternative and no excessive apology.

Prompt 10 — The clean ending: Ends a client relationship professionally, with no ambiguity about whether it’s continuing.

Plus: the pre-project “no” template to head off most of these situations before they start.

Read more

But it makes AI nearly useless for the thing I need most: finding out where my thinking is wrong before I act on it.

I’ve built a set of prompts I run specifically when I want the model to push back. Not to validate, not to help me execute, not to make my idea sound better. To tell me what’s wrong with it. What I’ve missed. What I’m assuming that might not be true.

These 7 prompts are what that looks like in practice.

The problem with asking AI “what do you think?”

When you ask AI for feedback on an idea, you mostly get a list of strengths followed by a polite section on “potential considerations.” The considerations are usually the least threatening version of the real objections.

That’s not feedback. That’s a compliment with a footnote.

Getting genuine pushback requires telling the model explicitly that agreement is not useful. Every prompt in this chain does that in a different way, for different kinds of thinking problems.

Prompt 1: The steel-manned attack

What it does: Asks the model to build the strongest possible case against your idea, not a weak version of the objection but the version that would most damage the idea if it turned out to be true.

When to use it: When you have an idea or plan you feel confident about. Confidence is exactly when you most need this. The more certain you feel, the more useful it is.

The prompt:

I’m going to describe an idea I believe in. Your job is not to evaluate it fairly. Your job is to build the strongest possible case against it. Find the version of the counterargument that would most damage this idea if it turned out to be true. Don’t hedge. Don’t balance it with positives. Just attack it. Here’s the idea: [DESCRIBE YOUR IDEA OR PLAN IN AS MUCH DETAIL AS YOU CAN]

How to use it:

1. Describe your idea in full, including why you think it’s right and what you’re planning to do with it

2. Read the attack without defending yourself on the first pass. Just take it in

3. Note which objection you most want to dismiss immediately. That’s usually the one worth sitting with

Example input: “I’m planning to go all-in on long-form Substack content and cut social media entirely. My reasoning is that social is a distraction, the algorithm controls your reach, and long-form builds a real audience that you own.”

What you’ll get: A full attack on the idea. For this example: the model might argue that long-form content without social distribution is hard to grow from scratch, that “owning your audience” only matters once you have one, and that cutting social entirely removes the discovery mechanism most Substack writers depend on. None of that means the idea is wrong. But it means you’ve stress-tested it before committing.

Advanced note: If the attack feels weak or obvious, tell the model: “That’s the surface objection. What’s the deeper structural problem with this idea?” The first attack is often the most predictable counterargument. The second one tends to be more useful.

Those two prompts cover the most common failure modes. The next five go after more specific problems.

Prompt 3 — The commitment audit: Makes the case for reversing a decision you’ve already made, ignoring sunk costs entirely, as if you were evaluating it fresh today with everything you now know.

Prompt 4 — The rationalisation detector: Looks at an argument you’re making and tells you whether it reads like reasoning toward a conclusion or reasoning from one, with the specific passage where the logic jumps flagged.

Prompt 5 — The “is this actually working?” audit: Takes the evidence you’re using to conclude something is working and asks whether the evidence actually supports that, or whether you’re reading ambiguous signals as confirmation.

Prompt 6 — The pre-publish argument check: Finds the weakest point in the central argument of a piece you’re about to publish, specifically the place a sharp critic would focus, not the most obvious objection.

Prompt 7 — The smartest critic in the room: Responds to your idea as the most credible, well-informed person in your hardest-to-satisfy audience would, quoting the specific parts they’d take issue with.

Plus: the challenge kit, a set of pre-built pushback prompts you can drop into any conversation without setup.

Read more

The fix isn’t learning to write better prompts. It’s learning to use AI to find the right question before you ask for the answer. The model is better at surfacing what you’re actually trying to solve than most people are at describing it cold.

These prompts run in sequence. By the end, you’ll have a question specific enough to get a genuinely useful answer, and a method you can repeat whenever you’re stuck at the fuzzy front end of a problem.

Step 1: Get the vague thing out of your head

Prompt 1: Problem dump processor

What it does: Takes an unformed, half-baked description of what you’re trying to figure out and turns it into a set of more specific sub-questions you might actually mean.

When to use it: When you know something is wrong or something needs solving but you can’t articulate what, exactly. The feeling of “I need to figure out X” with no clear sense of what X is.

The prompt:

I’m trying to figure something out but I can’t articulate it clearly yet. Here’s my messy version: [WRITE YOUR VAGUE PROBLEM IN PLAIN LANGUAGE, AS MESSY AS IT IS]. Don’t answer this yet. Instead, identify 5-7 more specific questions I might actually be trying to answer. For each one, write the question in one sentence and note in brackets whether it’s primarily a knowledge question, a decision question, or a diagnosis question.

How to use it:

1. Write your vague problem as you’d describe it to a friend who’d be patient with you. Don’t clean it up. Messy is better here

2. Paste it in place of [WRITE YOUR VAGUE PROBLEM IN PLAIN LANGUAGE, AS MESSY AS IT IS]

3. Read through the 5-7 questions and mark the one or two that make you think “yes, that’s closer to what I mean”

Example input: “I’m not sure if my newsletter is working. Like I’m writing it and people are reading it but I don’t know if it’s actually doing what I want it to do or if I should change something or if I’m just not being patient enough.”

What you’ll get: Five to seven specific questions pulled from the vague one. Something like: “Is your newsletter growing at a rate consistent with your goals? (decision)” and “Do you have a clear definition of what ‘working’ means for this newsletter? (diagnosis)” and “Are the right people subscribing, or just a lot of people? (diagnosis).” One of those will land closer than the others.

Advanced note: The question type in brackets matters. Knowledge questions (”what is X”) and decision questions (”should I do X”) need completely different prompts. If you try to answer a decision question with a knowledge answer, you’ll get information that doesn’t help you move. Sorting by type before you go further saves time.

That prompt does the thing most people try to do in their head and can’t: it breaks a vague feeling into specific questions you can actually work with.

The rest of the chain takes those questions further.

Prompt 2 — Assumption surfacer: Identifies every assumption baked into your chosen question, because a question with a false assumption in it produces a confident answer to the wrong thing.

Prompt 3 — Question sharpener: Adds the context, constraints, and decision criteria your question left out, then rewrites it as a fully-briefed prompt specific enough to get a genuinely useful answer.

Prompt 4 — Unknown-unknowns prompt: Surfaces the 3–4 things most people in your situation don’t know to ask about, that would significantly change the answer if they did.

Prompt 5 — Decision reframer: For decisions you’ve been stuck on, finds out whether the stuckness is about the options or about how the decision is framed, and whether you’re choosing between the real options at all.

Prompt 6 — Frame-breaker: Challenges the premise of your question entirely, for situations where the problem isn’t how you’re asking but what you’re asking about.

Plus: the question-building template you copy once and use on any problem where you’re not sure what you’re actually asking.

Read more

That’s not a consistency problem. It’s a signal problem. Authority content does something specific: it gives readers a reason to remember who said it. Calendar-filling content is correct and forgettable in equal measure.

These 7 prompts audit your existing content and tell you which side of that line you’re on.

Step 1: Work out what you’re actually saying across all your content

Prompt 1: Content position extractor

What it does: Reads a batch of your recent posts and identifies the underlying positions you’re taking, separate from the topics you’re covering.

When to use it: First. Before any other audit step. Topics and positions are different things, and most people conflate them. Writing about productivity is a topic. “Most productivity advice optimises for output when the real problem is decision load” is a position.

The prompt:

I’m going to paste several pieces of my recent content below. For each piece, identify: the topic (one phrase), the position (the specific claim or argument being made, in one sentence), and whether the position is something a reasonable person in this space could disagree with. If a piece doesn’t have a clear position, say so explicitly. Here are the pieces: [PASTE 5-10 RECENT POSTS OR ARTICLES]

How to use it:

1. Pull your last 5-10 published pieces. Substack posts, LinkedIn updates, threads, whatever you publish most regularly

2. Paste them together in place of [PASTE 5-10 RECENT POSTS OR ARTICLES]

3. Save the output. The “no clear position” flags are your first data point

Example input: Eight recent Substack posts across a range of topics the writer covers regularly.

What you’ll get: A table of topics vs. positions. Some pieces will have clear, arguable positions. Others will come back as “this piece describes X but doesn’t take a position on it.” That ratio tells you more about your content than any engagement metric.

Advanced note: The “could a reasonable person disagree” test matters. If the answer is no for most of your content, you’re not building authority, you’re confirming things people already believe. Authority comes from being right about something others haven’t said yet, or said clearly enough.

That prompt does something most content audits skip entirely: it separates topic from position, which is where the real diagnosis starts.

The next six prompts complete the audit and tell you what to do with what you find.

Prompt 2 — Theme concentration checker: Takes your position list and tells you whether your content is building toward a coherent point of view or spreading across unrelated claims that don’t add up to anything.

Prompt 3 — Authority signal detector: Scores each piece on four criteria that separate authority-building content from calendar filler, with specific evidence from the text for each score.

Prompt 4 — High-signal piece analyser: Takes your highest-scoring pieces and extracts the specific decisions that made them score higher, so you can repeat those decisions deliberately.

Prompt 5 — Buried claim excavator: Scans your archive for positions that were stated but not developed, where the seed of a strong authority claim got one sentence when it needed a full piece.

Prompt 6 — Pre-publication authority brief: Forces you to define your original claim, your evidence, and your memorable framing before you start drafting, so you’re not hoping a position emerges during writing.

Prompt 7 — Position stress tester: Takes the central claim of a piece you’re about to publish and tries to defeat it, so you know which counterarguments to address and which reveal a genuine limitation in the claim.

Plus: the content authority scorecard to track every piece going forward.

Read more

The problem isn’t the model. It’s the input. A style guide describes your voice from the outside. It lists adjectives (”conversational, direct, a bit dry”) and formatting rules and things to avoid. That’s not how voice works. Voice is what you do when nobody told you to do anything. It’s which words you reach for when you’re just thinking out loud.

These 8 prompts teach AI your voice by showing it patterns you didn’t know you had.

Step 1: Extract your patterns from writing you’ve already done

Prompt 1: Writing sample analyser

What it does: Reads a piece of your existing writing and identifies the specific linguistic patterns that make it sound like you, not general style advice.

When to use it: Before anything else. Pick your single best piece, the one that sounds most like you at your most natural.

The prompt:

I’m going to paste a piece of writing below. Analyse it for the specific patterns that make it distinctive. Don’t describe the tone in adjectives. Instead, identify: sentence length patterns (with rough averages), how the writer opens paragraphs, what they do at the end of paragraphs, any recurring grammatical quirks, how they handle transitions, whether they use rhetorical questions and how, how they signal emphasis without bold or caps, and any vocabulary they reach for repeatedly. Be specific. Quote examples from the text. Here’s the writing: [PASTE YOUR WRITING SAMPLE]

How to use it:

1. Pick one piece you’d hold up as “this sounds exactly like me” (a Substack post, a long LinkedIn post, an email you wrote when you weren’t trying to sound professional)

2. Paste it in place of [PASTE YOUR WRITING SAMPLE]

3. Save the full output. You’ll use it in Prompt 2

Example input: A 600-word Substack post the writer considers their most natural-sounding piece, written without a brief.

What you’ll get: A specific analysis. Not “uses a conversational tone” but “opens most paragraphs with a short declarative sentence under 10 words, often stating something the reader might disagree with.” That specificity is the whole point.

Advanced note: Run this on three pieces, not one. Pick writing from different formats if you can (a post, an email, a comment thread). Patterns that show up across all three are your actual voice. Patterns that appear in only one piece might be the topic talking, not you.

That prompt gives you something most AI voice work never produces: specific, quotable patterns rather than adjectives.

The rest of the chain turns that analysis into a voice profile you can use on anything. Subscribers get seven more prompts:

Prompt 2 — Cross-sample pattern matcher: Compares analyses from multiple writing samples and identifies only the patterns that appear consistently across all of them.

Prompt 3 — Voice profile builder: Turns your pattern analysis into a compact, reusable profile under 200 words, written as instructions rather than personality adjectives.

Prompt 4 — Format tone calibrator: Adapts your master profile for a specific format, because how you write a Substack post isn’t how you write a cold email.

Prompt 5 — Voice-led drafter: Writes a first draft with your voice profile baked into the generation instructions, not added as an afterthought.

Prompt 6 — Voice deviation checker: Reads a draft and flags every sentence that doesn’t match your profile, with a specific note on what’s wrong and a suggested fix.

Prompt 7 — Edit-based profile updater: Takes the edits you made to an AI draft and extracts new voice patterns from them, then updates your profile to reflect what you actually changed.

Prompt 8 — Voice stress tester: Generates a draft on a topic you’d never normally write about to reveal which parts of your profile are robust and which only work when the subject is familiar.

Plus: the fill-in-the-blank voice profile template you build once and reuse every time.

Read more

There’s a faster way. Pull the transcript, feed it to AI, and run it through a prompt chain that does the heavy lifting while you’re still finishing your coffee. By the time the video ends, you’ve got a working draft.

These 8 prompts do that. They go in sequence, and each one hands off directly to the next.

Step 1: Get the transcript into a usable state

Prompt 1: Raw transcript cleaner

What it does: Strips filler, timestamps, and speaker noise from a raw YouTube transcript so it’s actually readable.

When to use it: Right after you copy-paste a transcript out of YouTube or a tool like Tactiq.

The prompt:

I’m going to paste a raw YouTube transcript below. It contains timestamps, filler words, repetition, and speaker labels. Clean it up into readable prose. Don’t summarise it, don’t change the meaning, don’t add anything. Just remove the noise and fix the sentence flow. Keep every substantive idea. Here’s the transcript: [PASTE TRANSCRIPT]

How to use it:

1. Open the YouTube video, click the three dots under the title, select “Show transcript”

2. Copy everything and paste it in place of [PASTE TRANSCRIPT]

3. Run it and save the output as your working source text

Example input: A 20-minute video on cold email strategy. The raw transcript has timestamps every 30 seconds, “um” and “you know” throughout, and a few repeated sentences where the speaker restated a point.

What you’ll get: A clean, readable version of everything the speaker said, with nothing added or removed. About 30-40% shorter than the raw version, but complete.

Advanced note: If the transcript is longer than about 4,000 words, split it into two halves and run them separately. Paste both outputs together before moving to Prompt 2.

That prompt alone turns an unusable wall of timestamps and filler into something you can actually work from.

The rest of the chain takes that clean transcript and turns it into a published article.

Prompt 2 — Key insight extractor: Pulls the 8–12 most useful, specific ideas from the cleaned transcript as a structured list.

Prompt 3 — Article angle finder: Generates five possible angles from your insight list, each with a different hook and target reader.

Prompt 4 — Structured outline builder: Turns your chosen angle into a full outline with section headers, summaries, and the key point each section needs to land.

Prompt 5 — Section drafter: Writes one section at a time, using your outline and source insights to stay grounded in the video’s content.

Prompt 6 — Full draft tightener: Cuts padding, sharpens transitions, and flags any section where the central claim is unclear.

Prompt 7 — Hook writer: Three hook options using different entry points, written last when you know exactly what the article delivers.

Prompt 8 — Notes repurposer: Pulls three standalone Substack Notes from your finished article, each framed as its own observation.

Plus: the complete 8-prompt chain template you can copy and run on any video in under 30 minutes.

Read more

The problem isn’t the idea. It’s that you’re trying to turn raw instinct into a structured argument in real time, which nobody can do well. What you need is a private pressure test before you’re in front of anyone.

This 8-prompt sequence walks your idea through the same gauntlet a sceptical investor or executive would put it through. Do them in order. By the end you’ll have a written business case, a one-paragraph pitch, and a clear picture of where your idea is strong and where it isn’t.

Prompt 1: The plain-English summary

What it does: Forces you to state your idea without jargon or hedging, in terms a non-expert can understand.

When to use it: Before anything else. If you can’t complete this prompt clearly, the idea isn’t ready.

You are a plain-language editor. I have a business idea I want to describe clearly. Ask me nothing. Take what I give you and rewrite it as a single paragraph of no more than 100 words. Use simple words. Remove all jargon, filler and vague language. Every sentence must say something specific. Here is my idea: [DESCRIBE YOUR IDEA IN YOUR OWN WORDS, AS ROUGH AS YOU LIKE]

How to use it:

1. Paste the prompt and dump your idea in the placeholder. Don’t overthink it.

2. Read the output. If anything still feels vague, run it again with that section rewritten.

3. Save this paragraph. It becomes the opening of your business case.

Example input: My idea is a subscription service for small restaurants that gives them access to a shared fleet of delivery drivers, so they don’t have to pay Deliveroo’s commission on every order.

What you’ll get: A tight, specific summary you can read aloud in 20 seconds without stumbling.

Advanced note: If the output is too generic, add one sentence to your input that says who this is specifically for and what they currently do instead of your solution.

That prompt alone will stop you walking into a meeting with a vague idea dressed up as a plan.

But one paragraph isn’t a business case. Subscribers get seven more prompts that take it the rest of the way:

Prompt 2 — The problem statement: Articulates exactly who has this problem, what they do instead, and what it costs them.

Prompt 3 — The market size estimate: Builds a TAM/SAM breakdown from your own numbers, with every assumption flagged.

Prompt 4 — The assumption audit: Lists what your idea depends on being true, ranked by how badly each one kills you if it’s wrong.

Prompt 5 — The unit economics model: Shows what needs to be true for the business to make money, under three growth scenarios.

Prompt 6 — The competitive position: Forces an honest look at existing alternatives and what it would actually take to beat them.

Prompt 7 — The devil’s advocate session: Ten sharp questions a sceptical executive would ask, with your weakest answers flagged.

Prompt 8 — The business case document: Assembles everything into a structured, shareable document, approximately 1,000 words.

Plus: the 8-prompt progress tracker to keep all your outputs in one place.

Read more