Every tell the protocol names, one card each, with a bad example, a better one and the fix. “Zero allowed” means none in finished writing. A limit is a density cap. “Watch” is worth a second look but is not an error on its own. Agents: patterns.md · patterns.json.
Chatbot residue that no human ever types.
Internal tokens from ChatGPT, Grok and other tools that survive a copy-paste. Nothing else produces them.
BeforeRevenue grew in Q3 :contentReference[oaicite:2]{index=2}.
AfterRevenue grew in Q3 (company filing, 14 October).
Fix: Delete the token. If a real source sat behind it, cite that source by name.
Links copied out of a chatbot answer carry the chatbot in the URL.
Beforehttps://example.com/report?utm_source=chatgpt.com
Afterhttps://example.com/report
Fix: Strip the utm_source parameter.
Only a language model talks about its training data.
BeforeAs of my last update, the tool supports six languages.
AfterThe tool supported six languages in March.
Fix: Delete the whole sentence. If the information may be stale, say what date it comes from.
Assistant voice leaking into text that a person is supposed to have written.
BeforeAs an AI language model, I cannot give legal advice, but here are some points.
AfterThis is not legal advice. Three points to check with your lawyer:
Fix: Delete the whole sentence.
A bracketed blank means the text was generated and never read.
BeforeDear [Client Name], thank you for choosing [Company].
AfterDear Ms Ortega, thank you for choosing Harbor Print.
Fix: Stop and ask the user for the missing value. Never guess it.
The chatbot sign-off. A person who wrote the text does not offer to write more of it.
BeforeLet me know if you would like me to expand on any of these points!
After(nothing; the text ends on its last point)
Fix: Delete it. End on the last real point.
Describing the text instead of writing it. It reads like the wrapper around a chatbot answer.
BeforeCertainly! Here's a rewritten version that is more concise:
After(start with the text itself)
Fix: Delete it. The reader can see the structure.
Asterisks and hashes show up raw in email, LinkedIn and SMS. Only a model pastes them there.
BeforeHi Sam, **quick update:** ### Next steps
AfterHi Sam, quick update on next steps.
Fix: Use the channel's own formatting, or none.
Sentence shapes models reach for by default.
The loudest short-form tell of 2026. It invents a position nobody held so it can knock it down.
BeforeIt's not just a tool, it's a new way of working.
AfterThe tool cut our review time from two days to one.
Fix: State the claim directly, or write two plain sentences.
Three balanced items is what a model produces when it has one real example and fills to three.
BeforeOur platform is efficient, scalable, and reliable.
AfterThe platform handled 3,000 orders on Black Friday without a restart.
Fix: Keep the items you can defend. Two or four is fine. Uneven lists read as human.
Throat-clearing before the point. An editor deletes it first.
BeforeIn today's fast-paced digital world, businesses need to move quickly.
AfterOur competitors ship a release every week. We ship one a quarter.
Fix: Open on the actual claim.
An appeal to research that has no name, date or link behind it.
BeforeStudies have shown that remote teams are more productive.
AfterIn a 2015 Stanford experiment with a Chinese travel agency, staff working from home were 13% more productive.
Fix: Name the study, the number and the year, or delete the claim. Never attach a citation you cannot verify.
A sentence shaped like insight that carries no information. Delete it and nothing is lost.
BeforeAt the end of the day, the key is finding the right balance.
AfterWe cap meetings at two afternoons a week.
Fix: Delete it. If the paragraph now feels thin, it was thin before.
Paragraphs that open with Furthermore, Moreover, Additionally read like an outline being filled in.
BeforeMoreover, the new process reduces errors.
AfterThe new process also caught the invoice bug we missed in May.
Fix: Delete the connector. Let the order of the paragraphs carry the logic. Zero in anything under 400 words.
The five-paragraph-essay ending: challenges, then the future, then a restatement.
BeforeAs the landscape continues to evolve, one thing is clear: adaptability will be key.
AfterWe will know by March whether the pilot pays for itself.
Fix: End on the last real thing you have to say.
A participle bolted onto the end of a sentence to claim significance it has not earned.
BeforeThe team shipped the feature early, highlighting the importance of collaboration.
AfterThe team shipped the feature a week early.
Fix: Cut the tail, or make it its own sentence with a real subject.
Praise for the question instead of an answer to it.
BeforeGreat question! You're absolutely right to ask about pricing.
AfterPricing starts at $12 a seat.
Fix: Delete it. Start with the answer.
Models avoid plain "is" and "has" and reach for "serves as" and "boasts".
BeforeThe library serves as a hub for the community and boasts a large archive.
AfterThe library is where the town meets. It has 40,000 photographs of the harbour.
Fix: Use "is" or "has" unless the fancier verb adds meaning.
Everything is pivotal, vital and lasting. The reader stops believing any of it.
BeforeThe launch marks a pivotal moment and leaves an indelible mark on the industry.
AfterThe launch doubled our weekly sign-ups.
Fix: Say what happened and let the fact carry its own weight.
Brochure adjectives. They describe how the writer wants you to feel, not the thing.
BeforeNestled in the heart of the city, this breathtaking venue offers world-class service.
AfterThe venue is two streets from the station and seats 120.
Fix: Replace the adjective with the detail that made you reach for it.
Rotating synonyms so a noun never repeats. The reader wonders whether "the firm" and "the organisation" are different things.
BeforeAcme raised prices. The firm said the organisation needed to cover costs, and the Ohio-based company expects…
AfterAcme raised prices. Acme said it needed to cover costs and expects…
Fix: Pick one name and repeat it, or use a pronoun.
"From X to Y" where X and Y are not two ends of any scale. It sounds broad and says little.
BeforeWe help everyone from startups to governments with everything from pricing to culture.
AfterWe help software companies with 10 to 200 staff set their prices.
Fix: Name the things you mean, or drop the phrase.
A drumroll before the point. Social-media copy taught models to set up every fact like a punchline.
BeforeThe result? A 40% drop in errors. Here's the kicker: it took one afternoon.
AfterErrors dropped 40%, and the change took one afternoon.
Fix: State the result in a normal sentence.
Three clipped fragments in a row ("No fluff. No filler. Just results.") is a template, not a voice.
BeforeNo fluff. No filler. Just results.
AfterThe report is two pages and every number in it has a source.
Fix: Write one sentence that says the thing.
"Whether you're a X or a Y" tries to address everyone and so addresses no one.
BeforeWhether you're a seasoned pro or just starting out, this guide has you covered.
AfterThis guide is for people setting up their first online shop.
Fix: Write to the one reader this text is for.
Two hedges on one claim. Honest uncertainty names its limit; stacked hedges just blur it.
BeforeThis could potentially help to possibly reduce costs.
AfterThis should cut costs if the supplier holds its price.
Fix: Keep one hedge and say what it depends on.
Words whose frequency jumped after 2022.
Words like delve, underscore and showcase spiked in published text after ChatGPT launched. Two in one sentence is a fingerprint.
BeforeThis report delves into the intricate landscape of fintech, underscoring its multifaceted nature.
AfterThis report looks at how 40 fintech startups priced their first product.
Fix: delve → look at · underscore → show · tapestry → describe the thing · landscape → field · foster → build · harness → use · navigate → handle.
Latinate verbs standing in for plain ones.
BeforeWe leverage AI to streamline and facilitate onboarding.
AfterWe use AI to fill in the onboarding forms.
Fix: utilize/leverage → use · facilitate → help · streamline → simplify · commence → start · endeavour → try.
Phrases so worn they carry no meaning.
BeforeThis is a real game-changer that will move the needle.
AfterThis halves the time it takes to close a month.
Fix: Say what changed, what the easy part is, what moved.
Very, truly, incredibly: volume without information.
BeforeSales grew significantly and the team was incredibly happy.
AfterSales grew 34%.
Fix: Delete it, or replace it with a number.
Marks that give away the tool or the training data.
Humans use dashes too. The tell is density: well above 10 per 1,000 words, several per paragraph. Banning dashes outright is now its own tell.
BeforeThe plan — which we tested — worked — mostly.
AfterThe plan worked, mostly. We tested it for two weeks.
Fix: Keep under 15 per 1,000 words and one per paragraph in short pieces. Swap the rest for commas, colons, parentheses or full stops.
Spaced dashes ( — ) and tight dashes (word—word) in the same text suggest pasted fragments.
BeforeThe plan — tested twice—worked.
AfterThe plan, tested twice, worked.
Fix: Pick one convention and hold it.
A plain-text substitute for a dash that usually marks copied output.
BeforeWe shipped it -- finally.
AfterWe shipped it. Finally.
Fix: Use a full stop or a comma.
Typographic quotes in a code comment, SMS or form field are a copy-paste fingerprint.
BeforeReply “yes” to confirm.
AfterReply "yes" to confirm.
Fix: Use straight quotes when the destination is plain text. Keep curly quotes in typeset documents.
The rocket, lightbulb, sparkles and check-mark set is the signature of generated social posts.
Before🚀 Faster builds 💡 Smarter defaults ✅ Fewer bugs
AfterBuilds are faster, the defaults changed, and we closed 30 bugs.
Fix: At most one emoji per message, and only when it carries tone a word cannot.
Structure used because it is safe, not because the content needs it.
"**Speed:** the system is faster" on every bullet is the house style of chatbot answers.
Before- **Speed:** The system is faster. - **Cost:** It is cheaper.
AfterThe system is faster and costs less to run.
Fix: Write the bullet as a sentence, or turn the list into prose. Fine in reference documentation.
Every Word Capitalised in a body-copy heading is a default, not a choice.
Before## How To Improve Your Team Workflow
After## How to improve your team workflow
Fix: Sentence case: capital on the first word and on names only.
When everything is bold, nothing is.
BeforeThe **new plan** gives **every team** a **clear budget** and **faster approvals**.
AfterThe new plan gives every team its own budget. **Approvals now take one day.**
Fix: Keep bold for the one phrase a skimming reader must not miss.
Models reach for bullets because bullets are safe. Three sentences that follow from each other are not a list.
Before- The deadline moved - So we cut scope - So we shipped on time
AfterThe deadline moved, so we cut scope and still shipped on time.
Fix: Write the sentences.
A divider between every short section is decoration.
BeforeIntro --- Point one --- Point two
AfterIntro, then two sections under their own headings.
Fix: Let headings and white space separate sections.
Text that reads fine and says nothing.
"Many", "numerous", "a wide range of" stand where a number should be.
BeforeNumerous customers use a wide range of our features.
After212 customers use at least three of our features.
Fix: Use the number. If you do not have it, ask for it.
The most reliable signal. Strip every tell from an empty paragraph and it is still empty.
BeforeNutrition plays a crucial role in overall wellness.
AfterSwapping one soda a day for water cut my sugar intake by about 35 grams.
Fix: At least one number, name or date per 100 words. If the source has none, ask the user for one. Never invent it.
A sentence you can delete without losing information was filler.
BeforeCommunication is essential in any team. It helps everyone stay aligned.
AfterWe moved the stand-up to Slack and got back 40 minutes a day.
Fix: Delete it. Fewer than a third of sentences should survive deletion without loss.
The worst outcome of humanising. A made-up statistic is real harm; a generic sentence is only a style problem.
Before(an invented "73% of managers" to satisfy the specificity rule)
After"How many clients did you onboard last quarter? I will use your number."
Fix: Ask the user for the number. Use [brackets] only while drafting, and never ship them.
The tells that come from scrubbing too hard.
Choppy fragments, forced lowercase and no subordinate clauses: the style that replaced the em dash as a tell.
Beforewe tried it. it worked. kind of. not really.
AfterWe tried it for a month, and it worked about half the time.
Fix: Keep long sentences when the idea is long. Merge two short sentences into one.
Planted typos, fake "I think maybe" hedges and padded personal detail are now tells of their own.
Beforei think maybe this could work lol, teh numbers look ok
AfterThis could work. The numbers hold if churn stays under 3%.
Fix: Write correctly. Keep the author's real quirks; do not add new ones.
An author's lowercase starts or double-dot pauses are the fingerprint you are protecting.
Before(normalising a writer who always starts emails with "hey..")
After(leaving "hey.." exactly as they write it)
Fix: Keep the quirks from the source text, including ones that look like errors.
Swapping "leverage" for "use" inside the same balanced 20-word sentence changes nothing measurable.
BeforeWe use AI to simplify and help onboarding. (same shape as before)
AfterAI now fills in the onboarding forms. It took a week to set up.
Fix: Fix the structure first, the vocabulary second.
Wzorce typowe dla polskiego tekstu z AI.
Trzy zdania pod rząd zaczynające się tak samo to retoryczny szablon.
BeforeChcemy rosnąć. Chcemy zatrudniać. Chcemy zmieniać rynek.
AfterW tym roku chcemy zatrudnić dwie osoby do obsługi klienta.
Fix: Połącz dwa z nich albo zacznij inaczej.