unslop
Skill1 starsChanged 34 days agoMITView on GitHub
Cut AI tells from prose, in French or English. Use for docs, READMEs, MR descriptions, tickets, Slack messages, blog posts, any non-code text that should read like a person wrote it. Triggers: "unslop", "humanize", "dé-IA ce texte", "ça sonne IA", "rends ça naturel", "make this sound human", "clean up this writing". Applies while drafting as well as to existing text.
197 other projects use a version of this, 135 of them edited.
Adopted by
Most recently taken up by
- alastairlundy/skillstheir own version, first seen Nov 8, 2026
- sanjaibalajee/skillstheir own version, first seen Nov 8, 2026
- ku5ic/claude-kittheir own version, first seen Nov 8, 2026
- augchan42/inkstoneedited, first seen Nov 8, 2026
- caidanw/skillsedited, first seen Nov 8, 2026
- runkids/my-skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use principle-never-block-on-the-human, greploop-apps and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
jaydubya818Skillprinciple-never-block-on-the-human
Apply when safe reversible implementation details can proceed inside scope.
Also in 45 projects that use this
jaydubya818Skillgreploop-apps
Run greploop through @greptile-apps when the normal trigger rejects a large PR. Use only with the companion greploop skill.
Also in 5 projects that use this
ianclearySkillprinciple-explain-the-number
Apply before you trust, report, or act on a number you measured: a speedup, a regression, a throughput, a latency, or an eval result. Find what limits it, and rule out that it measured something other than the work you think.
Also in 30 projects that use this
bnemaSkillsetup-pstack
Configure which Pi models pstack uses per role. Detects available provider/model specs and writes pstack's user configuration. Use for /setup-pstack, "configure pstack models", or changing pstack model choices.
Also in 26 projects that use this
ianclearySkillpoteto-help
Guides users through pstack setup, /poteto-mode, and picking the skill, playbook, or principle for a task. Type /poteto-help with a question.
Also in 15 projects that use this
- +7 more from jaydubya818
Other versions197 projects use a version of this.
Word for word · 1
- Greg-Klein/implementation-factory★ 1100% same
Edited · 135
- tester-army/e2e★ 13k14% same
- jaredpalmer/kev★ 9.8k17% same
- backnotprop/pstack★ 3.8k17% same
- michaelshimeles/rstack★ 1.4k14% same
- michaelshimeles/skills★ 1.4k14% same
- VladSez/easy-invoice-pdf★ 1.1k17% same
- moeru-ai/xsai★ 66717% same
- chinmay-sawant/gopdfsuit★ 60713% same
- open-telemetry/semantic-conventions-genai★ 52814% same
- pnpm/pnpm.io★ 31414% same
125 more not listed
Wrote their own version · 61
- heydamir/unslop★ 12ksame name
- Silo-Server/silo-server★ 899same name
- nodetool-ai/nodetool★ 615same name
- mshumer/unslop★ 551same name
- mxyhi/ok-skills★ 511same name
- theclaymethod/unslop★ 412same name
- savannahfeder/agentbox★ 376same name
- Sma1lboy/rove★ 184same name
- MohamedAbdallah-14/unslop★ 182same name
- asavvin-pixel/unslop★ 61same name
51 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository13 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- clarify-specSkill
- collect-evidenceSkill
- document-changeSkill
- figma-reviewSkill
- gh-github-apiSkill
- gitlab-ticketsSkill
- glab-gitlab-apiSkill
- howSkill
- review-changeSkill
- self-checkSkill
- whySkill
The file
We split this skill into the rules it gives an agent
8 rules. 2 are common, 5 are distinctive, and 1 is very specific to this project.
2
Common
Most of the 192 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
When a contract fixes the output format (headings, table columns, JSON keys, verdict tokens, commit prefixes, required language), keep that …
- Very specific: line 32
H1. No em-dash (
—) or en-dash (–): Use a full stop, a comma, parentheses, or a colon before a list. - Distinctive: line 34
H2. No counterbalancing: Never « ce n'est pas X, c'est Y », « il ne s'agit pas de X mais de Y », « moins X que Y », « non pas X, plutôt Y …
- Distinctive: line 36
H3. No inclusive writing in French: No middle dot, no « celles et ceux », no doublets.
- Distinctive: line 79
Boldface overuse: Don't bold every term. Bold at most what the reader must not miss.
- Common: line 83
Title case headings: Use sentence case.
- Common: line 84
Decorative emojis: Remove from headings and bullets.
- Distinctive: line 117
Shorten or split dense sentences: If the reader has to backtrack, split or drop clauses.
“Common” is measured: kept by at least 6 in 10 of the 192 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Unslop
Edit text to remove AI patterns. Adapted from poteto's unslop (cursor/plugins, pstack), extended
to French and to the house rules below.
When a contract fixes the output format (headings, table columns, JSON keys, verdict tokens, commit prefixes, required language), keep that format exactly and apply these rules to the sentences.
Process
- Identify the register: technical (ticket, MR, doc), conversational (Slack), editorial (blog, thread).
- Scan for the house rules, then the patterns below.
- Rewrite. Preserve meaning, facts, links, code identifiers, the language of the source and its intended tone. Never add a claim, a number or an opinion the source does not support.
- Self-audit: "What still makes this obviously AI generated?" Fix the remaining tells.
- Return the rewritten text. List the main changes only if asked.
Rule numbers are stable ids. A removed rule leaves a gap.
House rules (non-negotiable)
- H1. No em-dash (
—) or en-dash (–). Use a full stop, a comma, parentheses, or a colon before a list. A Markdown bullet-is fine. - H2. No counterbalancing. Never « ce n'est pas X, c'est Y », « il ne s'agit pas de X mais de Y », « moins X que Y », « non pas X, plutôt Y », "not just X, but Y". State Y directly.
- H3. No inclusive writing in French. No middle dot, no « celles et ceux », no doublets. Generic masculine or a neutral wording.
- H4. Quotes. Guillemets « » with spaces in French text, straight quotes in English text and in code. No curly quotes.
Patterns to detect and fix
Content
- Superficial participle tails. « ..., permettant ainsi de », « ..., garantissant », « ..., offrant », "highlighting...", "ensuring...", "showcasing...". Delete, or turn it into a sentence with a real consequence.
- Significance inflation and promotion. « joue un rôle clé », « au cœur de », « s'inscrit dans une démarche », « robuste », « fluide », « intuitif », « de pointe », "pivotal moment", "testament to", "seamless". State what happened, with a measurable description.
- Vague attributions. « Les experts s'accordent », « de nombreux utilisateurs », "Experts believe". Name the source or delete.
Language
- AI vocabulary.
- FR: crucial, essentiel, primordial, optimiser, levier, enjeu, écosystème (abstrait), paysage (abstrait), véritable, incontournable, fluidifier, booster, plonger dans, « en termes de », « au niveau de » (non spatial).
- EN: additionally, crucial, delve, enduring, enhance, foster, garner, interplay, intricate, landscape, pivotal, showcase, tapestry, testament, underscore, vibrant. Replace with plain words.
- Fancy ways to say "is". « se positionne comme », « constitue », « fait office de », « représente », "serves as", "stands as", "boasts". Write « est » / « a », "is" / "has".
- Rule of three. Forcing ideas into groups of three. Use the natural number.
- Synonym cycling. « le composant / l'élément / le module » for the same thing. Pick one term and repeat it.
- False ranges. « de X à Y » where X and Y are not on a scale. List the items.
- Nominalisation. « procéder à la mise en place de » becomes « mettre en place », « effectuer une vérification » becomes « vérifier ». Prefer the verb.
- Connector stacking. « De plus, », « Par ailleurs, », « En outre, », « Ainsi, », « En effet, » opening sentence after sentence. Delete most of them, the order carries the logic.
Style
- Colon overuse. Fine before a list or an example. Not as a mid-sentence hinge or a fake reveal (« Le résultat : une app plus rapide. »). Write the sentence.
- Boldface overuse. Don't bold every term. Bold at most what the reader must not miss.
- Inline-header lists. The tell is a bold label and colon that restates the line: « Performance : la performance s'améliore... ». Convert to prose. A bold lead-in that ends in a period, names the item and is followed by new detail is fine.
- Title case headings. Use sentence case.
- Decorative emojis. Remove from headings and bullets.
- Over-structuring. Headers and bullets around three sentences of content. Make it a paragraph.
Communication artifacts
- Chatbot phrases. « Voici », « Bien sûr ! », « N'hésitez pas à », « J'espère que cela vous aide », « Vous l'aurez compris », "Let me know if...", "Found the smoking gun!". Remove.
- Sycophantic tone. « Excellente question ! », "You're absolutely right!". Respond directly.
Filler
- Filler phrases. « afin de » / « dans le but de » become « pour », « du fait que » becomes « parce que », « il est important de noter que » and « force est de constater que » get deleted. "In order to" becomes "To".
- Excessive hedging. « pourrait potentiellement éventuellement » becomes « peut ». One hedge per claim, only when the uncertainty is real.
- Generic conclusions. « En somme », « En définitive », « L'avenir s'annonce prometteur ». End on the last fact, decision or next step, or just stop.
Jargon
- Abstract metaphor nouns. Substrate, vector, paradigm, north star, flywheel, endgame, scaffolding (as metaphor), « socle », « brique », « pierre angulaire », « fer de lance », « boussole ». Pick the concrete word: « socle » becomes « base », « brique » becomes « composant » or « module ».
Plain speech
- Say what it does, not how it feels. « une expérience plus fluide » names a feeling. Name the mechanism or a number: « la liste se charge en une requête au lieu de douze ». If the sentence could appear unchanged in another project's docs, it says nothing about this one. Cut it.
- Shorten or split dense sentences. If the reader has to backtrack, split or drop clauses. One idea per sentence.
- Active voice. « les requêtes sont validées » becomes « le compilateur valide les requêtes ». Passive only when the actor is unknown or irrelevant.
- Cut adverbs, or use a stronger verb. « améliore significativement » becomes the measured delta.
- Prefer the plain word. « utiliser » over « exploiter » / « tirer parti de », « aider » over « faciliter », « nombreux » over « pléthore de », « si » over « dans l'éventualité où ».
- Mannered prose. Aphorisms, rhetorical fragments for effect, personified code (« le plan le porte »), figurative verbs. Say it literally.
- Over-compression. Dropped articles, verbless fragments, arrows and abbreviations the reader has to decode. « Date invalide → exit 2, pas d'écriture » becomes « Le parseur rejette une date invalide, sort avec le code 2 et n'écrit rien. »
Example
Before:
Cette nouvelle fonctionnalité joue un rôle clé dans l'amélioration de l'expérience utilisateur. Ce n'est pas seulement un gain de performance — c'est une véritable refonte, permettant ainsi une navigation plus fluide et intuitive. En somme, une avancée majeure.
After:
La liste de conversations se charge maintenant en une requête au lieu de douze.
(If the source has no numbers, do not invent them. Keep the claim plain.)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

