unslop: text that leaves the session sounds like a person

Two passes over a text someone else will read: the register (sound like a person, in this language, for this relationship and channel) and the named-tell audit (no marks of a machine, tell by tell). Run it on anything about to leave the session: a README, a post, an email, a landing page, a commit body. How the harness talks to the user is orient, not this.

Find the source without ceremony

A bare "bro", "unslop", "write like a human", "hazlo más natural" points to the assistant's last answer. If the user pasted text, work on that. If they asked for new copy, draft it directly. Ask for text only when no source exists.

The register pass

1. Preserve the payload. Names, facts, numbers, links, commitments, constraints, uncertainty and necessary warnings stay. Never invent evidence, opinions, anecdotes or certainty to make it flow.
2. Match the room. The user's language, fitted to audience, relationship and channel. A natural legal notice is still precise; a natural Slack message can be loose. Human is not always casual.
3. Cut the bot-shaped scaffolding. Throat-clearing, needless summaries, repeated conclusions, corporate filler, fake enthusiasm, headings or bullets that do not help the reader navigate.
4. Restore a real cadence. Concrete verbs, direct sentences, natural contractions. The words a person in this context would choose, without manufactured slang, typos or choppy fragments.

Aim	The result
Plain, not simplistic	Easy to follow without losing the idea
Concise, not incomplete	No padding; every load-bearing fact remains
Warm, not performative	Appropriate care without canned empathy
Specific, not decorated	Concrete nouns and verbs instead of hype

When a text needs both passes: audit first, register last.

The pass

1. Scan against the tables below. Every hit gets located, not sensed. If you cannot name the tell, it is not a finding.
2. Rewrite in place. Preserve meaning, facts, numbers, links, commitments and the intended tone.
3. Restore the voice (next section). A text stripped of tells and nothing else reads as sterile, which is its own tell.
4. Self-audit once more. Ask: what still makes this obviously machine written? Fix that too.
5. Return the text. On request, return the list of named hits alongside it.

Removing tells is half the job

Voiceless writing is as obvious as slopped writing. After the cuts, put something back:

• Have an opinion. React to the facts instead of listing balanced pros and cons.
• Vary the rhythm. Short sentence. Then a longer one that takes its time and earns the length.
• Admit complexity. "Impressive and slightly unsettling" beats "impressive".
• Use "I" when it fits. First person is not unprofessional.
• Allow some mess. Perfect symmetry looks manufactured.
• Be specific. Not "this is concerning" but "it churns for six hours at 3am and nobody is watching".

Content tells

Tell	Fix
Puffery: "pivotal moment", "testament to", "evolving landscape", "indelible mark"	State what happened
Name dropping outlets with no context	Pick one, quote what it said
Decorative -ing clauses: "highlighting...", "ensuring...", "showcasing..."	Delete, or expand into a real claim with a source
Promotional adjectives: "nestled", "vibrant", "breathtaking", "renowned", "must-visit"	Neutral description
Vague attribution: "experts believe", "reports suggest", "critics argue"	Name the source or cut the claim
The formulaic challenge arc: "despite challenges, X continues to thrive"	Specific facts instead of the arc

Language tells

Tell	Fix
AI vocabulary: additionally, crucial, delve, enhance, foster, garner, interplay, intricate, landscape, pivotal, showcase, tapestry, testament, underscore, vibrant	The plain word
Fancy ways to say "is": "serves as", "stands as", "boasts", "features"	"is", "has"
"Not just X, but Y"	State the point directly
Rule of three: forcing every list into three items	The number the facts have
Synonym cycling: protagonist, main character, central figure, hero in one paragraph	Pick one and repeat it
False ranges: "from X to Y" where X and Y share no scale	List the items
Abstract metaphor nouns: substrate, wedge, vector, locus, nexus, primitive, surface, bedrock, scaffolding, paradigm, flywheel, north star, gold-plating, ratchet, evacuate	The concrete word: base, add, way, more than the job needs, move out

Style and punctuation tells

Tell	Fix
Em dashes, anywhere	End the sentence, or use a comma. Swapping in parentheses trades one tell for another
Colons as mid-sentence connectors	Let the point stand as its own sentence. Colons are for lists and examples
Bold on every proper noun and acronym	Bold what the reader must not miss, nothing else
Inline-header lists whose bold label restates the line: "Performance: performance improved"	Prose. A bold lead-in that ends in a period and is followed by genuinely new detail is fine
Title Case Headings	Sentence case
Decorative emoji in headings and bullets	Remove
Curly quotes	Straight quotes

Filler, hedging and plain speech

Tell	Fix
"In order to", "due to the fact that", "it is important to note that"	"To", "because", delete
Stacked hedges: "could potentially possibly be argued that it might"	"may", or the claim itself
Generic conclusion: "the future looks bright"	A specific plan, date or number
Chatbot phrases: "I hope this helps", "let me know if", "certainly", "great question"	Remove
Cutoff disclaimers: "while specific details are limited"	Find the detail or drop the sentence
Sycophancy: "you're absolutely right"	Answer directly
Feeling instead of mechanism: "the database stays close at hand"	The mechanism or the number: ".toSQL() returns the exact string sent to the database"
A sentence that would fit unchanged in another project's docs	Cut it. It says nothing about this one
Dense sentences the reader has to reparse	One idea per sentence
Passive with a hidden actor: "queries are validated"	Name the actor: "the compiler validates queries"
Adverbs propping up weak verbs: "runs quickly", "significantly improves"	A stronger verb or the measured number
Fancy synonyms: utilize, leverage, facilitate, numerous, in the event that	use, use, help, many, if

Anti-patterns

Anti-pattern	Why it fails	Do this instead
Report the tells and leave the text unfixed	The user asked for a clean text, not a diagnosis	Fix in place, list on request
Cut every tell and stop	Sterile prose is the other AI tell	Restore opinion, rhythm and specifics
Rewrite the meaning while cleaning the style	The edit silently changes the claim	Facts, numbers, links and commitments stay exact
Strip a term because it sounds technical	Precision lost to style is a defect	Keep the term, explain it once
Touch code, commands, URLs, quotes or regulated wording	Those are not prose	Leave them byte for byte
Claim the result will pass an AI detector	Nobody can promise that	Promise named tells removed, nothing more
Add typos, slang or filler to seem human	It performs humanity and sounds less credible	Match the actual relationship and channel
Use the same casual voice for every audience	A client email, a README and a condolence need different registers	Infer the room before choosing the cadence
Announce and justify every edit	The preamble is the friction the user asked to remove	Lead with the finished words
Flatten a voice the author chose on purpose	Some authors do write long, warm or formal	Match the intended tone, cut only the machine marks

Output boundary

Return the cleaned text as an artifact, with no preface. Keep harness commentary, the findings list and any compass block clearly outside it, so nothing that is not the text can be copied into an email or a document by accident.

Where this ends

• orient owns how the harness talks to the user: STE-style, short, every answer stands alone. This skill is for text the user sends to someone else.
• brand-voice owns the standing voice system across writers and channels.
• technical-writing owns structure: what belongs in a tutorial, a reference, a README.
• landing-copy owns conversion architecture. Clean its words only after it has chosen them.