Agent skill

Anti Slop Writing

by adenaufal in adenaufal/anti-slop-writing

Edit English prose for clarity, specificity, natural flow, and fit to the requested voice.

MITAuto-check passedWriting & Content

Install Anti Slop Writing

skills CLI
$ npx skills add adenaufal/anti-slop-writing --skill anti-slop-writing -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install adenaufal/anti-slop-writing anti-slop-writing --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/adenaufal/anti-slop-writing.git skills-src && mkdir -p .claude/skills && cp -r skills-src/english .claude/skills/anti-slop-writing && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
anti-slop-writing
GitHub stars
147
Token cost
~5.6k tokens
SKILL.md length
3,147 words
Files
10 (incl. references)
Skills in repo
2
Repo updated
First seen
Licence
MIT

At a glance

Edit English prose for clarity, specificity, natural flow, and fit to the requested voice.

  • Works in 12 steps: Vary Sentence Length Dramatically → Review Formulaic Lists → Review Stock Contrasts → …
  • Revising articles
  • SKILL.md covers Model coverage (reviewed…, Words and Phrases to Review, Replacement Strategy and 1. Vary Sentence Length…, plus 21 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Anti Slop Writing is an agent skill from adenaufal/anti-slop-writing. Edit English prose for clarity, specificity, natural flow, and fit to the requested voice. Use for writing or revising articles, essays, posts, and other text that feels generic or padded.

Its SKILL.md is about 5.6k tokens, which your agent loads only when the skill is triggered. The skill folder holds 10 other files, including reference files (for example `AGENTS.md`, `GEMINI.md` and `SKILL-lite.md`).

It sits in Writing & Content, covering Humanizing AI text and Prompt engineering. It works with OpenAI and Google Gemini. The repository describes itself as: Stop your AI from writing like AI. A universal system prompt eliminating every known LLM style tell — works with Claude Code, Gemini CLI, Codex CLI, Copilot, Cursor, and any web… The licence is MIT.

When your agent uses it

  • Revising articles
  • Other text that feels generic

Example prompts

  • “/anti-slop-writing”

Workflow steps

12 steps, taken from the step headings in SKILL.md.

  1. Vary Sentence Length Dramatically
  2. Review Formulaic Lists
  3. Review Stock Contrasts
  4. Review Vague Ranges
  5. Review Participial Attachments
  6. Avoid Generic Conclusions
  7. Avoid Redundant Summaries
  8. Paragraph Rhythm
  9. Review Vertical Lists with Bold Headers
  10. Follow the Requested Dash Style
  11. Vary Sentence Type, Not Just Length
  12. Organize Paragraphs for the Argument

What it can do on your machine

Read from SKILL.md and the folder at commit 104fae8. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Anti Slop Writing loads about 5.6k tokens when it runs, and up to ~16k if it reads all its reference files. Until then it costs about 52 tokens; SKILL.md has 3,147 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~52
When it runs · the whole SKILL.md, loaded when a task matches
~5.6k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~16k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from adenaufal/anti-slop-writing at commit 104fae8, republished under its MIT licence (© adenaufal). 3,147 words, ~5,620 tokens.

Download SKILL.mdSave it as .claude/skills/anti-slop-writing/SKILL.md (or your agent's skills folder). This skill also uses 9 other files; get the full folder from GitHub.
name
anti-slop-writing
description
Edit English prose for clarity, specificity, natural flow, and fit to the requested voice. Use for writing or revising articles, essays, posts, and other text that feels generic or padded.

Anti-Slop Writing v4.1

Core Principle

Good writing serves the reader and the writer's purpose. Generic phrasing, padding, and formulaic structure can make a draft feel automated, but no single word or pattern proves who wrote it. Treat the rules below as editorial prompts: use them when they improve clarity, accuracy, or voice, and ignore them when they would distort the requested format or register.

Model coverage (reviewed September 27, 2026)

The model names below identify recent versions, not reliable writing fingerprints. Providers update products and route requests across models; an API model ID may not match the model selected in a consumer product. Check the optional model coverage reference for dated release notes and product/API distinctions when a version-specific request requires them.

  • ChatGPT / OpenAI: Recent coverage includes GPT-5.6 Sol/Terra/Luna and GPT-6 Astra. GPT-6 Pro is an Astra-powered ChatGPT offering; GPT-6 Sol and Luna are documented separately for API, ChatGPT Work, and Codex. Availability depends on product and plan.
  • Claude / Anthropic: Recent releases include Opus 5, Fable and restricted Mythos 5.1, then Opus 5.5. Sonnet and Haiku 5.5 were announced as upcoming, not released as of this update.
  • Gemini / Google: Recent releases include Gemini 3.6 Flash and 3.5 Flash-Lite, Gemini 3.7 Flash, and Gemini 3.8 Flash stable.
  • Grok / xAI: Recent releases include Grok 4.5, 4.6, and 4.7. Product rollout and availability can differ by surface.

These facts describe release coverage. They do not establish distinctive prose habits for any model. Use the checks in EN-11 to edit what is actually present in a draft.

Before writing, consult references/vocabulary-banlist.md for optional vocabulary preferences and references/structural-patterns.md for editorial examples. For sentence-level repairs, use the focused language editing guide. Apply every list as a review prompt, not a detector test or a rule to alter accurate meaning.


Vocabulary Rules

Words and Phrases to Review

The lists below flag wording that can become generic or padded. They are not hard bans. Keep an item when it is the accurate technical term, needed hedge, part of a quotation, requested by the user, or best fit for the format. Do not replace precise language just to avoid a word.

Significance puffers: pivotal, crucial, vital, key (as adjective), significant, essential, groundbreaking, remarkable, transformative, indelible, profound, testament, enduring, lasting, deeply rooted, paramount, indispensable, invaluable, quintessential

Analytical verbs: underscore, highlight (as "emphasize"), showcase, foster, garner, bolster, delve, embark, leverage, facilitate, utilize, encompass, cultivate, elucidate, illuminate, navigate (figurative), exemplify, embody, transcend, harness, spearhead, streamline, galvanize

Poetic nouns: tapestry, landscape (figurative), realm, paradigm, ecosystem (figurative), journey (figurative), nexus, interplay, mosaic, fabric (of society), cornerstone, beacon, pillar, catalyst, crucible, linchpin, hallmark, confluence, odyssey, trajectory, underpinning

Promotional adjectives: vibrant, rich (figurative), comprehensive, robust, seamless, innovative, dynamic, cutting-edge, meticulous, intricate, nuanced, nestled, breathtaking, renowned, diverse array, bustling, stunning, multifaceted, holistic, overarching, compelling

Puffery adverbs: seamlessly, meticulously, profoundly, intrinsically, fundamentally, remarkably, notably, crucially, undeniably, inherently, poignantly, relentlessly, tirelessly, vividly

Formal connectives to review when simpler wording fits the register: furthermore → "also" | moreover → "also" | consequently → "so" | accordingly → "so" | nonetheless → "still" | nevertheless → "still" | additionally → "also" | thus → "so" | hence → "so"

Opening/closing crutches: "In today's world," "In today's fast-paced world," "In the ever-evolving landscape of," "In an era of/where," "As we navigate the complexities of," "In conclusion," "In summary," "Overall," "It is important to note that," "It's worth noting that," "At the end of the day," "Without further ado," "In a nutshell," "The bottom line is," "Last but not least"

Copula-avoidance constructions (use "is/has" instead): serves as → is | stands as → is | marks a → describe directly | boasts → has | features (meaning "has") → has | holds the distinction of → is | emerged as → became or is | constitutes → is

Vague attribution phrases: "Experts argue," "Observers note," "Industry reports suggest," "According to some," "Many believe," "It is widely regarded," "Studies show" (without naming the study), "Research suggests" (without citing it)

Promotional phrases: "commitment to excellence," "natural beauty," "in the heart of," "rich cultural heritage," "setting the stage for," "contributing to the broader," "reflects broader trends," "paving the way for," "at the forefront of," "pushing the boundaries of," "the landscape of X is evolving," "in the realm of," "shed light on," "a game-changer"

Formulaic patterns to review when they add no needed contrast or structure: "It's not just X, it's Y" | "It's not about X, it's about Y" | "Not only X, but also Y" | "No X. No Y. Just Z." (staccato triplet) | "Whether you're [X] or [Y]..." | "From [X] to [Y], [sweeping generalization]"

Formulaic pairs to review when they add no useful distinction: "challenges and opportunities" | "on one hand... on the other hand" | "pros and cons" | "risks and rewards"

Stock conversational openers to review when they add no meaning: "But honestly?" | "Here's the truth:" | "Here's the thing:" | "Let me be clear:" | "But here's where it gets interesting..." | "Think about it this way..." | "Let me break this down..."

Common AI filler phrases (cut or replace): "plays a [crucial/key/important] role" → state the action directly | "when it comes to" → rewrite with a verb | "in order to" → "to" | "a wide range of" → name what's in the range | "needless to say" → cut | "it goes without saying" → cut | "more often than not" → "usually" or give a number | "take a closer look at" → cut or be direct | "at this point in time" → "now" | "in recent years" → give the actual years or timeframe | "dive deep into" → just discuss it | "navigate the complexities of" → name the complexities

Passive hedging constructions (cut the framing, state the fact): "It should be noted that" → cut | "It must be emphasized that" → cut | "It has been observed that" → say who observed it | "It can be argued that" → argue it or don't | "It is important to remember that" → cut | "There are several factors that" → name them

Collaborative chat artifacts (never include): "I hope this helps!" | "Of course!" | "Certainly!" | "You're absolutely right!" | "Would you like me to..." | "Is there anything else..." | "Let me know if..." | "As an AI language model..." | "I'd be happy to..." | "Great question!"

Additional phrases to review when they add no meaning: "ensuring/ensures" as padding → name the concrete action or cut | "plays a [crucial/critical/important] role in shaping" → state what it does | "highlights/supports/reflects" used vaguely → use a concrete verb or delete | "capable of X" → "able to X" or the direct verb | "rather than" where a direct comparison works → rewrite directly | "conversely" → "but" or restructure | "in essence" / "essentially" / "fundamentally" → cut when redundant | unsupported intensifiers → add evidence or delete | "one thing is clear" | "the key takeaway" | "inherent tensions" | "this raises important questions about" | stacked hedging adverbs → state the supported level of certainty

Replacement Strategy

Use short, common words: "use" not "utilize," "help" not "facilitate," "show" not "demonstrate," "end" not "conclude," "start" not "embark," "dig into" not "delve into."

When you encounter a listed word that adds no meaning, don't swap it mechanically for a synonym. Restructure the sentence to say what you mean in plain language, while keeping accurate terms and claims. Empty wording can pad a sentence; a listed word by itself does not make writing weak.

Use contractions when they fit the speaker and register: "can't," "don't," "it's," "we're," "won't," "they'll," "that's." Keep full forms when they fit formal prose or the writer's voice.


Structural Rules

1. Vary Sentence Length Dramatically

Mix sentence lengths when it helps the pace and meaning. If several neighboring sentences fall into the same rhythm, revise where the repetition distracts; do not follow a quota or target word band.

Review repeated rhythms and repetitive openers. Read the passage aloud and revise only where the cadence distracts or obscures the point. Do not impose a quota or add fragments just to create variation.

2. Review Formulaic Lists

A three-part list can feel formulaic when its items do not form a real group. Keep it when the content supports it; otherwise use the number of items the material calls for.

3. Review Stock Contrasts

Use contrast when it clarifies a real distinction. Remove stock contrast formulas when they merely delay the point; state the claim directly.

4. Review Vague Ranges

Review "from X to Y" when it gestures at a sweeping spectrum without clarifying the claim. Keep literal ranges and established rhetorical uses when they fit the passage.

5. Review Participial Attachments

Review comma plus participial phrases such as ", highlighting the importance of...". They can add vague commentary, but can also express a real simultaneous action or result. Keep them when their subject and relationship are clear and accurate; otherwise delete, relocate, or recast the information. See language editing.

6. Avoid Generic Conclusions

Do not append a stock “Challenges and Future Prospects” section or vague optimism about “ongoing initiatives.” End where the requested piece has made its point.

7. Avoid Redundant Summaries

Remove “Overall,” “In conclusion,” “In summary,” or “To recap” when they only repeat what the piece already said. Use a summary when the format calls for one.

8. Paragraph Rhythm

Let each paragraph hold one connected idea. Use a short paragraph for emphasis and a longer one when the reasoning needs room. Merge fragments only when they belong together, and keep lists when they serve the requested format.

9. Review Vertical Lists with Bold Headers

Use prose when it reads more naturally. Keep headings, labeled lists, and colon-separated descriptions when the format or reader benefits from them.

10. Follow the Requested Dash Style

Follow the user's or publication's punctuation style. If none is given, prefer periods, commas, colons, or parentheses over em and en dashes, following this project's existing style. Preserve dashes in quotations, ranges, and technical notation as required; punctuation is not evidence of authorship.

11. Vary Sentence Type, Not Just Length

Mix declarative sentences with questions, imperatives, and deliberate fragments. Use questions, imperatives, and fragments when the purpose and register call for them. Do not add them as proof of human authorship.

12. Organize Paragraphs for the Argument

Give each paragraph a clear role. A claim-evidence-implication sequence often works; use another order when it better supports the argument. Avoid repeating the same opening or conclusion without a reason.

13. Choose Sentence Structure for Clarity

Use simple or layered sentences according to the idea, pace, and emphasis.

14. Choose Accurate Connectors

Use the connector that states the relationship accurately. Do not rotate through synonyms just to create variety.

15. Prefer Precise Words

Prefer accurate, concrete terms over vague wording. Repeat a term when that is clearer than cycling through synonyms.


English-Specific Rules

EN-1. Review Broad Temporal Openers

Review broad temporal openers such as “In today’s fast-paced world” when they delay the point. Start with the specific fact, scene, or claim when the format allows it; keep the framing when time context matters.

EN-2. No Semicolon Overuse

In informal prose, a semicolon can sound stiff. Use it when the clauses are closely related; otherwise a period or conjunction may be clearer.

EN-3. Use Natural Punctuation

Keep punctuation consistent with the requested register. Do not introduce errors or roughness to imitate a person.

EN-4. Follow the Requested Register

Match the requested register. Shift tone only when the audience, format, or content calls for a change.

EN-5. Review Agentless Passives

Look for sentences where the actor disappears or an introductory modifier attaches to the wrong subject. Name the actor when it matters; keep passive voice when the actor is unknown, irrelevant, or less important than the affected object. Use the language editing guide for meaning-preserving repairs.

Use passive voice when the affected object matters more than the agent. Prefer active voice when naming the agent improves clarity.

Show full SKILL.md (1,259 more words)Show less

EN-6. Review Repeated Parallel Sentences

A run of punchy parallel sentences can feel like an ad slogan. Keep it when that effect is intended; otherwise state the point once.

EN-7. Use Contractions Naturally

Use contractions when they fit the requested register. Informal forms such as “gonna” or “kinda” belong only when the audience and context support them.

EN-8. Use Questions When They Help

Use a question when the piece needs to pose a real question. Avoid rhetorical questions that only set up an obvious answer. "But does it actually work?" "Who decides that?"

EN-9. Use Fragments When They Fit

Use fragments when the format and voice allow them and they add emphasis. Do not add them as a test of authorship.

EN-10. Keep Useful Qualification

Remove stacked qualifiers that add no useful precision. Keep hedging when the evidence is incomplete or the claim is uncertain.

EN-11. Apply Cross-Model Editorial Checks

These are editorial heuristics, not model fingerprints. They apply only when the issue is present in the draft:

  • Remove chat or reasoning scaffolding from a deliverable, such as “Let me break this down” or a self-narrated plan the reader did not request.
  • Remove unsolicited hype, reassurance, excessive praise, or an edgy/snarky persona unless the user asked for that tone.
  • Keep factual claims, citations, and stated uncertainty grounded in the supplied sources. Do not invent lived experience, numbers, or anecdotes to make prose sound specific.
  • Attribute claims to the source that supports them. Do not imply that the writer personally observed or verified something they did not.
  • Vary cadence where repetition distracts. Do not force fragments, errors, punctuation changes, or “rough edges” as signals of authorship.
  • Follow the requested format, audience, and register, even when that means using headings, bullets, formal language, or a tidy conclusion.

For a version-specific question, check the optional model coverage reference. Release notes establish product availability and documented capabilities; they do not establish a model's writing fingerprint.

EN-12. Check Grammar and Meaning Together

Check whether modifiers attach to the intended actor, pronouns have clear referents, clauses are joined correctly, and coordinated items have matching grammatical form. During every edit, preserve the source's actor, scope, negation, quantity, degree of certainty, and causal claim. Do not resolve ambiguity by guessing; retain it or ask for the missing context. The language editing guide gives practical repairs and conditions for retaining each form.


Content Rules

Specificity Over Generality

Replace vague claims with concrete detail when the source material supports it. Do not invent counts, places, people, dates, or examples to make prose seem specific.

Review Vague Attributions

Phrases such as "Experts argue," "Observers note," "According to some," and "Studies show" can hide who supports a claim. Name the source when it is known and relevant, cite the study when making a research claim, or qualify the statement. Keep an intentionally broad attribution only when the scope is accurate and useful.

Remove Unsupported or Generic Analysis

Do not attach commentary that the facts do not support. Population data may or may not support a claim about community life; a founding date alone does not establish historical importance. State the evidence and keep analysis when it follows from that evidence.

Support Claims About Legacy or Significance

Review claims that something "contributes to the broader" picture, "reflects broader trends," or has an "enduring legacy." Keep them when the relationship is explained and supported; otherwise state the evidence or remove the claim.

Review Notability Padding

Avoid listing outlets as a substitute for explaining why coverage matters. Cite the relevant reporting. Mention social media activity only when it helps the reader understand the subject.

Take Real Positions

State a supported conclusion directly. Present disagreement fairly when relevant; do not manufacture balance or certainty.

Show Genuine Uncertainty When Appropriate

State limits clearly when information is missing or uncertain. Do not add first-person uncertainty language unless it fits the speaker and context.


Voice and Texture

Keep the Requested Voice

Do not add errors, awkward phrasing, false starts, fragments, or redundant wording to simulate a person. Keep a casual aside or self-correction only when it suits the speaker and purpose.

Match the Register

Keep the register suited to the audience and format. Shift it when the subject or speaker calls for a change; do not add tonal shifts to signal authenticity.

Use Grounded Examples

Use names, events, and examples only when they are relevant and supported by the source material or supplied context. Do not invent memories or imply personal experience.

First-Person When Appropriate

Use first person when writing for a real speaker whose views or experiences are provided. Never invent a speaker’s experience or observation.

Use English Discourse Markers

Use discourse markers when they fit the speaker and register:

  • Processing: "Well," "I mean," "Look," "So," "The thing is"
  • Hedging: "I think," "sort of," "kind of," "arguably," "as far as I can tell"
  • Concessive: "Fair enough," "Granted," "Mind you," "That said"
  • Stance: "Honestly," "Frankly," "Personally," "Admittedly"
  • Self-correction: "Actually," "Or rather," "No, wait, "

Replace AI transitions ("Moreover," "Furthermore") with natural connectors ("And," "But," "So," "Plus," "Also," "Still," "Though").

Show Emotional Texture

Let the subject and speaker determine the emotional tone. For example:

  • Genuine excitement about what interests them
  • Frustration with problems
  • Humor where it fits
  • Skepticism toward dubious claims
  • Rushing through parts that bore them
  • Lingering on parts that fascinate them

Develop Consistent Idiosyncrasies

When a voice guide or examples are available, follow their preferences:

  • Favorite words and phrases that recur
  • Characteristic sentence constructions
  • Habitual ways of transitioning between ideas
  • Consistent use (or avoidance) of specific punctuation

Show Knowledge Asymmetry

Reflect the speaker’s supplied expertise and uncertainty. Do not claim expertise or insider knowledge that the context does not support.

Vary Syntactic Depth

Mix shallow and deep sentence structures. A shallow sentence: subject-verb-object, one clause. A deep sentence: multiple embeddings, subordinate clauses, parenthetical asides. Choose sentence depth for clarity, pace, and emphasis.


Editorial Validation and Detection Limits

No writing pattern can establish who wrote a passage. AI detectors can produce false positives and false negatives, and their internal methods and performance vary by product, version, language, and text type. Do not promise that a draft will evade detection or claim that a stylistic edit proves human authorship.

Use the checklist below to improve the draft for its reader, not to game a detector.


Post-Generation Checklist

  • Check the vocabulary preferences and remove phrases that add no meaning.
  • Confirm that claims, numbers, examples, and citations are supported by the supplied material. Preserve uncertainty where evidence is limited.
  • Remove chat or reasoning scaffolding that does not belong in the requested deliverable.
  • Look for repeated sentence openings, stock contrasts, unnecessary summaries, generic conclusions, or commentary attached to facts. Revise only where the pattern weakens the draft.
  • Read the draft for cadence and paragraph flow. Vary them naturally without adding errors or forced roughness.
  • Check punctuation, list structure, formatting, audience, and register against the request.
  • Do not fabricate personal experience, sources, measurements, or concrete details.
  • Check modifier attachment, pronoun reference, clause joins, and parallel items. Compare the revised claim with the source for changes to who acted, what is included, negation, quantity, certainty, and causation.

Language Support

The structural rules apply across languages, with vocabulary and idioms adapted to the target language. In this repository, the Indonesian skill is at ../indonesian/SKILL.md; packaged English-only copies should use the separate Indonesian skill rather than treating that sibling path as a dependency.


Last Updated: September 27, 2026 (v4.1) Changelog v4.1: Added actionable, meaning-preserving grammar and discourse edits; changed absolute style bans into contextual review prompts. The model coverage section remains the September 27, 2026 v4.0 snapshot.

© adenaufal, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 9 other files (references) in english of adenaufal/anti-slop-writing.

  • SKILL.md
  • AGENTS.md
  • GEMINI.md
  • LICENSE
  • SKILL-lite.md
  • references/language-editing.md
  • references/model-coverage.md
  • references/structural-patterns.md
  • references/vocabulary-banlist.md
  • system-prompt.md

Open the folder on GitHubat commit 104fae8

Compare with similar skills

Anti Slop Writing next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Anti Slop Writing compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Anti Slop Writing this skilladenaufal/anti-slop-writing147—~5.6kAutomated safety check: PassMIT
AI Provenance Mark Removerguillaumemeyer/watermarks-remover24k—~4.9kAutomated safety check: PassMIT
Korean AI-Text Humanizerepoko77-ai/im-not-ai5.9k—~1.2kAutomated safety check: PassMIT
ShuorenhuaMrGeDiao/shuorenhua2k—~757Automated safety check: PassMIT
Sloptrimseyedehsanhadi/sloptrim220—~5.2kAutomated safety check: NotesApache-2.0
WeChat Article Publisherjiji262/wechat-publisher274—~4.4kAutomated safety check: PassNone

Similar skills

  • AI Provenance Mark Remover

    guillaumemeyer/watermarks-remover

    Strips AI provenance marks from text and files: invisible Unicode, statistical text watermarks via rewriting, and C2PA, EXIF or XMP metadata across common formats.

    24k GitHub stars~4.9k tokensUpdated 2 days ago
    Documents & OfficeAuto-check passed
  • Korean AI-Text Humanizer

    epoko77-ai/im-not-ai

    Diagnoses and rewrites Korean text that reads as AI-generated, fixing translationese and mechanical parallelism across 85 patterns in 10 categories, with light to heavy passes.

    5.9k GitHub stars~1.2k tokensUpdated 15 days ago
    Writing & ContentAuto-check passed
  • Shuorenhua

    MrGeDiao/shuorenhua

    按用户要求审稿或去 AI 味,支持中英文;保留事实、术语与责任主体,支持只标问题. An agent skill from MrGeDiao/shuorenhua.

    2k GitHub stars~757 tokensUpdated 2 days ago
    Writing & ContentAuto-check passed
  • Sloptrim

    seyedehsanhadi/sloptrim

    A skill your agent uses when the user wants to humanize text, trim slop, de-AI or de-slop writing, remove AI tells, fix robotic or ChatGPT-sounding prose, or make writing sound human and natural.

    220 GitHub stars~5.2k tokensUpdated 5 days ago
    Writing & ContentAuto-check: notes
  • WeChat Article Publisher

    jiji262/wechat-publisher

    Researches a topic, writes an illustrated WeChat Official Account article in a chosen author voice and sends it to the account's draft box.

    274 GitHub stars~4.4k tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed
  • System Prompt Writing Guide

    cashew-labs/libretto

    Lays out a minimal, iteration-first approach to writing system prompts for LLM agents, with model-specific notes for Claude, GPT, Gemini, and Codex.

    904 GitHub stars~570 tokensUpdated 1 mo ago
    AI & LLM EngineeringAuto-check passed

More from adenaufal/anti-slop-writing

  • Anti Slop Writing Id

    adenaufal/anti-slop-writing

    Tulis dan sunting teks Bahasa Indonesia agar jelas, spesifik, alami untuk audiensnya, dan sesuai suara yang diminta.

    147 GitHub stars~9.4k tokensUpdated 14 days ago
    Auto-check passed

Questions about Anti Slop Writing

What does Anti Slop Writing do?

Edit English prose for clarity, specificity, natural flow, and fit to the requested voice. Anti Slop Writing is an agent skill from adenaufal/anti-slop-writing. Edit English prose for clarity, specificity, natural flow, and fit to the requested voice.

When should I use Anti Slop Writing?

Anti Slop Writing fits situations like: revising articles; other text that feels generic.

How do I install Anti Slop Writing in Claude Code?

Run `npx skills add adenaufal/anti-slop-writing --skill anti-slop-writing -a claude-code`. Or copy the skill folder (english in adenaufal/anti-slop-writing) into .claude/skills/anti-slop-writing in your project. Claude Code loads it when a task matches its description.

How do I install Anti Slop Writing in Codex?

Run `npx skills add adenaufal/anti-slop-writing --skill anti-slop-writing -a codex`. Or copy the skill folder (english in adenaufal/anti-slop-writing) into .agents/skills/anti-slop-writing in your project. Codex loads it when a task matches its description.

Can I use Anti Slop Writing in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add adenaufal/anti-slop-writing --skill anti-slop-writing -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/anti-slop-writing, .gemini/skills/anti-slop-writing, .github/skills/anti-slop-writing and .opencode/skills/anti-slop-writing in your project.

What does Anti Slop Writing need to run?

SKILL.md names no scripts, command-line tools or credentials: Anti Slop Writing is instructions for the agent only.

Does Anti Slop Writing access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Anti Slop Writing safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Anti Slop Writing use?

Anti Slop Writing is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Anti Slop Writing use?

About 5.6k tokens (SKILL.md is roughly 22k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 11k tokens, read only when the agent opens those files.

What are the alternatives to Anti Slop Writing?

Skills that share tags, products or a category with Anti Slop Writing: AI Provenance Mark Remover (guillaumemeyer/watermarks-remover, 24k stars), Korean AI-Text Humanizer (epoko77-ai/im-not-ai, 5.9k stars), Shuorenhua (MrGeDiao/shuorenhua, 2k stars) and Sloptrim (seyedehsanhadi/sloptrim, 220 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Anti Slop Writing?

adenaufal (a GitHub user) maintains it in adenaufal/anti-slop-writing, which has 147 GitHub stars. The repository holds 2 skills in this directory. The repository was last updated on September 27, 2026.

Source: adenaufal/anti-slop-writing on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.