Agent skill

AI Writing Detox

by jamditis in jamditis/claude-skills-journalism

Eliminates AI-generated writing patterns that erode reader trust.

MITAuto-check passedWriting & Content

Install AI Writing Detox

skills CLI
$ npx skills add jamditis/claude-skills-journalism --skill ai-writing-detox -a claude-code

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

GitHub CLI
$ gh skill install jamditis/claude-skills-journalism ai-writing-detox --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/jamditis/claude-skills-journalism.git skills-src && mkdir -p .claude/skills && cp -r skills-src/journalism-core/skills/ai-writing-detox .claude/skills/ai-writing-detox && 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
ai-writing-detox
GitHub stars
416
Token cost
~2.2k tokens
SKILL.md length
1,137 words
Files
2
Skills in repo
53
Repo updated
First seen
Licence
MIT

At a glance

Eliminates AI-generated writing patterns that erode reader trust.

  • Works in 7 steps: "delve" → delete or replace → "landscape" → delete or replace → "crucial" → is it actually crucial? → …
  • Editing articles
  • SKILL.md covers When to use, Banned words (delete or replace), Banned phrases and Banned structures, plus 7 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

AI Writing Detox is an agent skill from jamditis/claude-skills-journalism. Eliminates AI-generated writing patterns that erode reader trust. Use when writing or editing articles, docs, or press releases.

Its SKILL.md is about 2.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).

It sits in Writing & Content, covering Humanizing AI text and Copywriting. The repository describes itself as: Claude Code skills for journalism, media, and academia - verification, FOIA, data journalism, academic writing, and more. The licence is MIT.

When your agent uses it

  • Editing articles
  • Tasks that involve Humanizing AI text
  • Tasks that involve Copywriting

Example prompts

  • “Use the ai-writing-detox skill to eliminate AI-generated writing patterns that erode reader trust”
  • “/ai-writing-detox”

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. "delve" → delete or replace
  2. "landscape" → delete or replace
  3. "crucial" → is it actually crucial?
  4. "robust" → what does this mean here?
  5. "leverage" → just say "use"
  6. "comprehensive" → is this word doing any work?
  7. Starting "So," → delete

What it can do on your machine

Read from SKILL.md and the folder at commit e3e2172. 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

AI Writing Detox loads about 2.2k tokens when it runs. Until then it costs about 36 tokens; SKILL.md has 1,137 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~36
When it runs · the whole SKILL.md, loaded when a task matches
~2.2k

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 jamditis/claude-skills-journalism at commit e3e2172, republished under its MIT licence (© jamditis). 1,137 words, ~2,166 tokens.

Download SKILL.mdSave it as .claude/skills/ai-writing-detox/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
ai-writing-detox
description
Eliminates AI-generated writing patterns that erode reader trust. Use when writing or editing articles, docs, or press releases.

AI writing detox

Good writing is invisible. If readers notice the writing style, it's distracting from the content. AI patterns are noticeable, they break trust.

When to use

  • Writing or editing any journalism content
  • Creating press releases or media advisories
  • Drafting documentation or reports
  • Writing social media posts
  • Reviewing any AI-assisted text before publication

Banned words (delete or replace)

WordWhy it's badAlternative
delveAI signature wordexplore, examine, look at
realmPretentiousarea, field
tapestryPurple prose(delete entirely)
landscapeCorporate speaksituation, environment
leverageJargonuse
utilizeJargonuse
robustMeaninglessstrong, reliable
seamlessAlmost always falsesmooth, easy
comprehensiveRarely necessaryfull, complete
cutting-edgeClichémodern, new
holisticCorporatecomplete, full
synergyCorporate(delete, rewrite)
paradigmOverusedmodel, approach
empowerVague(be specific about what capability)
innovativeEmpty praise(describe what's new)
transformativeHyperbolicchanged, improved
sophisticatedVague filleradvanced, or describe what's complex
leveragingCorporate jargonusing
ecosystemOverused metaphorsystem, environment
rich(as modifier)(delete or be specific)
over (for quantity)AP rulemore than

Banned phrases

Throat-clearing (delete entirely)
  • "It's important to note that..."
  • "In today's [X] landscape..."
  • "Let's dive/delve into..."
  • "Without further ado..."
  • "In this article, we will..."
  • "It's worth mentioning that..."
  • "It goes without saying..."
  • "As we all know..."
Empty hedges
  • "To be fair..." / "To be honest..."
  • "At the end of the day..."
  • "When it comes to..."
  • "In terms of..."
  • "With respect to..."
AI enthusiasm
  • "This is a game-changer"
  • "...and that's a good thing!"
  • "Here's the thing:"
  • "...and that's okay!"
  • "...and I'm here for it"
Corporate buzzwords
  • "Moving forward..."
  • "Going forward..."
  • "At this point in time..."
  • "Due to the fact that..."
  • "In order to..."

Banned structures

Don't start sentences with:
  • "So," or "Well,"
  • "Now," (when not about time)
  • "Look," or "Listen,"
  • "Basically," or "Essentially,"
Don't end sentences with:
  • "...right?"
  • "...you know?"
  • "...if you will"

The substitution table

If you wroteWrite instead
utilizeuse
facilitatehelp
implementbuild, add, create
leverageuse
functionalityfeature
methodologymethod
in order toto
due to the fact thatbecause
at this point in timenow
a large number ofmany
in the event thatif
prior tobefore
subsequent toafter
in close proximity tonear
has the ability tocan

Case sensitivity

Always use sentence case for headings, not title case.

WrongRight
Getting Started With Your ProjectGetting started with your project
How To Use The ToolHow to use the tool
Best Practices For JournalismBest practices for journalism

Journalism-specific patterns to avoid

"Not just X—it's Y" pattern

Avoid: "This isn't just a news story—it's a wake-up call" Avoid: "This wasn't just a press conference—it was a turning point"

Do instead: State the thing directly without dramatic framing

"Fundamentally transforms" pattern

Avoid: "This fundamentally transforms how newsrooms operate" Avoid: "This represents a fundamental shift in journalism"

Do instead: Describe the actual change without hyperbole

Inflated claims in reporting

Avoid: "A bombshell revelation" Avoid: "A major milestone" Avoid: "A groundbreaking investigation"

Do instead: Let the facts speak. Readers judge significance.

Empty transitions

Avoid: "With that in mind..." Avoid: "Building on this foundation..." Avoid: "Taking this a step further..."

Do instead: Just make the next point

"It's not about X, it's about Y" pattern

Avoid: "It's not about the technology, it's about the people" Avoid: "This isn't about clicks. It's about trust."

Do instead: State the actual point without rhetorical setup. The X-not-Y construction is the dominant 2025-2026 ChatGPT/Claude rhetorical signature.

Em-dash overuse between independent clauses

Avoid: "The vote was close — only three council members opposed it — but the measure passed." Avoid: "Reporters arrived early — the press conference was already underway."

Do instead: Use commas, periods, or semicolons. Em-dashes are fine in moderation; reflexive use across every paragraph is a tell.

"The reality is..." / "The truth is..." openers

Avoid: "The reality is, most newsrooms can't afford this." Avoid: "The truth is, readers don't trust anonymous sources."

Do instead: Just state the claim. The opener adds words and signals AI-generated framing.

Tricolon abuse (reflexive three-item lists)

Avoid: "Clear, concise, and compelling reporting." Avoid: "Trust is built through accuracy, transparency, and consistency."

Do instead: Use one or two specific points. Three-item parallel lists are an AI rhythm pattern that reads as filler when used reflexively.

Show full SKILL.md (437 more words)Show less

Before and after examples

Example 1: General prose

AI slop:

In today's rapidly evolving media landscape, it's crucial to understand the multifaceted nature of AI tools. Let's delve into how these robust solutions can help journalists leverage cutting-edge technology.

Human writing:

AI tools do three things well: drafting, research, and analysis. Here's when to use each.

Example 2: News lede

AI slop:

The city council meeting was a comprehensive and transformative discussion that fundamentally addressed key issues affecting residents in a meaningful way.

Human writing:

The city council voted 5-2 Tuesday to raise property taxes by 3 percent.

Example 3: Feature intro

AI slop:

In the rich tapestry of American journalism, few stories are as compelling as the one about to unfold. This isn't just a profile—it's a journey into the heart of what makes local news truly matter.

Human writing:

Maria Rodriguez has published a newspaper for 47 years. She's never missed an issue.

Example 4: Analysis piece

AI slop:

Moving forward, stakeholders must leverage innovative approaches to navigate the challenging landscape of digital transformation in the journalism ecosystem.

Human writing:

Local papers that want to survive need to find digital revenue. Most haven't.

The verbal tic test

Read your text aloud. Does it sound like:

  • A TED talk introduction? Rewrite.
  • A LinkedIn post? Rewrite.
  • A press release? Rewrite.
  • Corporate communications? Rewrite.
  • How you'd explain it to a colleague? Keep it.

Quick self-check before publishing

Search your text for:

  1. "delve" → delete or replace
  2. "landscape" → delete or replace
  3. "crucial" → is it actually crucial?
  4. "robust" → what does this mean here?
  5. "leverage" → just say "use"
  6. "comprehensive" → is this word doing any work?
  7. Starting "So," → delete

Found any? Your writing needs another pass.

Red flags in technical/analysis writing

  1. Lists of near-synonyms: "comprehensive, sophisticated, and robust" (pick one or none)
  2. Excessive hedging: "may potentially be able to possibly..."
  3. Noun stacking: "production-ready deployment system infrastructure"
  4. Passive voice hiding agency: "It was determined that..." (by whom?)
  5. Circular definitions: "The system enables users to use the functionality"

The ultimate test

Before accepting any AI-assisted text, ask:

  1. Can I delete this word/phrase without losing meaning? → Delete it
  2. Is this the simplest way to say this? → Simplify
  3. Would I say this out loud to a colleague? → If not, rewrite
  4. Does this add information or just sound impressive? → If the latter, cut it

If readers think "AI wrote this," you've already lost their trust.


Good writing is invisible. If readers notice the writing, it's getting in the way of the content.

See also: hooks/ai-slop-detector.md for the automated companion to this skill.

© jamditis, 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 1 other file in journalism-core/skills/ai-writing-detox of jamditis/claude-skills-journalism.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit e3e2172

Compare with similar skills

AI Writing Detox 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.

AI Writing Detox compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
AI Writing Detox this skilljamditis/claude-skills-journalism416—~2.2kAutomated safety check: PassMIT
AI Copywritermikiarlo3/ai-copywriter1.2k—~12kAutomated safety check: PassMIT
Anti Vibe Writingweijt606/anti-vibe-writing120—~3.9kAutomated safety check: PassMIT
Copy Editingcoreyhaines31/marketingskills54k1 repos~4.5kAutomated safety check: PassMIT
Antislop Copywritingmiqdadbadjuber/anti-slop5.7k—~6.3kAutomated safety check: PassMIT
Humanize Koreansangrokjung/claude-forge852—~1.3kAutomated safety check: PassMIT

Similar skills

  • AI Copywriter

    mikiarlo3/ai-copywriter

    Write copy that converts and doesn't sound like a robot. An agent skill from mikiarlo3/ai-copywriter.

    1.2k GitHub stars~12k tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed
  • Anti Vibe Writing

    weijt606/anti-vibe-writing

    A skill your agent uses when the user wants to humanize, de-AI, polish, or warm up an AI-generated document, in English or Chinese.

    120 GitHub stars~3.9k tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed
  • Copy Editing

    coreyhaines31/marketingskills

    When the user wants to edit, review, or improve existing marketing copy, or refresh outdated content.

    54k GitHub starsUsed in 1 repo~4.5k tokens
    Writing & ContentAuto-check passed
  • Antislop Copywriting

    miqdadbadjuber/anti-slop

    Copy and text skill for antislop. An agent skill from miqdadbadjuber/anti-slop.

    5.7k GitHub stars~6.3k tokensUpdated 6 days ago
    Writing & ContentAuto-check passed
  • Humanize Korean

    sangrokjung/claude-forge

    Clean up translation-ese and AI-writing tics in Korean prose so it reads like a person wrote it — content (facts, numbers, proper nouns, quotes) stays 100% unchanged, only style and rhythm are…

    852 GitHub stars~1.3k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed
  • Orwell Writing

    tamdogood/builder-essential-skills

    A skill your agent uses when an agent is asked to draft, rewrite, edit, review, polish, copyedit, simplify, humanize, or create written prose, including creative writing, essays, posts, scripts…

    221 GitHub stars~1.1k tokensUpdated 1 mo ago
    Writing & ContentAuto-check passed

More from jamditis/claude-skills-journalism

All 53 skills in this repo
  • Web Design Picker

    jamditis/claude-skills-journalism

    A skill your agent uses when creating distinct website directions, a client review picker, asset catalog, previews, and Cloudflare-ready handoffs.

    416 GitHub stars~3.1k tokensUpdated 6 days ago
    Auto-check passed
  • Okf Wiki

    jamditis/claude-skills-journalism

    Builds an Open Knowledge Format (OKF) knowledge base from existing docs, notes, or a repo.

    416 GitHub stars~4.7k tokensUpdated 6 days ago
    Auto-check passed
  • Private Secret Scanning

    jamditis/claude-skills-journalism

    Local Gitleaks scans for staged changes, push ranges, and full history in private repos, with redacted reports.

    416 GitHub stars~1.8k tokensUpdated 6 days ago
    Auto-check passed
  • Data Journalism

    jamditis/claude-skills-journalism

    Acquire, clean, analyze, verify, visualize, and explain data for journalism.

    416 GitHub stars~1.6k tokensUpdated 6 days ago
    Auto-check passed
  • Document Design

    jamditis/claude-skills-journalism

    Creates print-ready HTML that exports to PDF. An agent skill from jamditis/claude-skills-journalism.

    416 GitHub stars~1.9k tokensUpdated 6 days ago
    Auto-check passed
  • Using Superjawn

    jamditis/claude-skills-journalism

    Establishes how to find and use skills, requiring Skill tool invocation before any response.

    416 GitHub stars~1.5k tokensUpdated 6 days ago
    Auto-check passed

Questions about AI Writing Detox

What does AI Writing Detox do?

Eliminates AI-generated writing patterns that erode reader trust. AI Writing Detox is an agent skill from jamditis/claude-skills-journalism. Eliminates AI-generated writing patterns that erode reader trust.

When should I use AI Writing Detox?

AI Writing Detox fits situations like: editing articles; tasks that involve Humanizing AI text; tasks that involve Copywriting.

How do I install AI Writing Detox in Claude Code?

Run `npx skills add jamditis/claude-skills-journalism --skill ai-writing-detox -a claude-code`. Or copy the skill folder (journalism-core/skills/ai-writing-detox in jamditis/claude-skills-journalism) into .claude/skills/ai-writing-detox in your project. Claude Code loads it when a task matches its description.

How do I install AI Writing Detox in Codex?

Run `npx skills add jamditis/claude-skills-journalism --skill ai-writing-detox -a codex`. Or copy the skill folder (journalism-core/skills/ai-writing-detox in jamditis/claude-skills-journalism) into .agents/skills/ai-writing-detox in your project. Codex loads it when a task matches its description.

Can I use AI Writing Detox 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 jamditis/claude-skills-journalism --skill ai-writing-detox -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/ai-writing-detox, .gemini/skills/ai-writing-detox, .github/skills/ai-writing-detox and .opencode/skills/ai-writing-detox in your project.

What does AI Writing Detox need to run?

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

Does AI Writing Detox 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 AI Writing Detox 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 AI Writing Detox use?

AI Writing Detox is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does AI Writing Detox use?

About 2.2k tokens (SKILL.md is roughly 8.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to AI Writing Detox?

Skills that share tags, products or a category with AI Writing Detox: AI Copywriter (mikiarlo3/ai-copywriter, 1.2k stars), Anti Vibe Writing (weijt606/anti-vibe-writing, 120 stars), Copy Editing (coreyhaines31/marketingskills, 54k stars) and Antislop Copywriting (miqdadbadjuber/anti-slop, 5.7k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains AI Writing Detox?

jamditis (a GitHub user) maintains it in jamditis/claude-skills-journalism, which has 416 GitHub stars. The repository holds 53 skills in this directory. The repository was last updated on October 4, 2026.

Source: jamditis/claude-skills-journalism on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.