Skip to content

Avoid AI Writing skill

Audit or edit English prose with context-aware patterns, protected source details and bounded passes.

Community · Conor Bronsdon
AuthorConor Bronsdon
LicenceMIT
Sourceconorbronsdon/avoid-ai-writing · SKILL.md on GitHub
Works inClaude Code · Codex, Gemini CLI, Cursor, Copilot
Read from the source on

avoid-ai-writing audits English prose for formulaic wording and rewrites only the justified, authorized spans.

It separates rewrite, detect-only and edit-in-place modes, preserves protected material and limits editing to two passes. Its patterns assess writing quality, not authorship.

From the skill's SKILL.md: # Avoid AI Writing — Audit & Rewrite

This is a community skill by Conor Bronsdon, not checked by us or by Anthropic. Skills can tell the agent to run code, read files or call the network. Before installing, read its SKILL.md and every script in the folder, and look for anything that does not match what it says it does.

What it does

  1. Read the pattern catalog and apply context exceptions.
  2. Preserve facts, quotes, code, uncertainty and the author’s voice.
  3. Run available local checks and report the actual pass count and residuals.

What you get

  • One final rewrite, a detect-only report or scoped file edits.
  • Verification notes distinguishing executed checks from model-only review.

When to use it

  • Clean up an English draft without inventing supporting details.
  • Audit boilerplate while leaving the original text unchanged.

When not to use it

  • Using style signals as proof of cheating, authorship or detector evasion.

What it needs

  • An agent that can read the bundled pattern references.
  • Node.js 18 or later for local detector and preservation scripts; model-only operation must disclose unavailable checks.

Files in the skill

File or folderWhat it is for
SKILL.mdScope, fidelity, modes and pass limits.
references/patterns.mdRequired pattern catalog and context exceptions.
detector/validate.jsMechanical source-preservation checks.
scripts/normalize-quotes.jsQuote convention normalization for editable prose.

Licence as found in the repository on 2026-11-17: MIT.

Example prompts

Type one of these after installing; naming the skill makes sure it is used.

  • Audit this English draft in detect-only mode; preserve intentional technical wording.
  • Edit only the introduction in this file, keeping facts, quotes and uncertainty intact.
  • Rewrite this announcement in one pass and report which local checks actually ran.

How to install avoid-ai-writing

Keep the support files

Copy the repository folder with references, detector and scripts; copying only SKILL.md loses required context and checks.

No separate writing API is required. Model access has the costs and privacy terms of the host agent. Local style signals do not establish who wrote the text.

A skill is a folder with a SKILL.md file. You install it once; the agent reads its short description at the start of each session and loads the full instructions only when a request matches. Pick the place you use:

Claude Code: copy the folder into your skills folder

Put the skill's folder in ~/.claude/skills/ to use it in every project on this computer, or in .claude/skills/ inside a repository to share it with everyone who works there (commit it).

git clone --depth 1 https://github.com/conorbronsdon/avoid-ai-writing.git
mkdir -p ~/.claude/skills/avoid-ai-writing
git -C avoid-ai-writing archive HEAD | tar -x -C ~/.claude/skills/avoid-ai-writing

Claude Code notices new or changed skills in these folders during the session. If the skills folder did not exist when the session started, run /reload-skills once. Then type /avoid-ai-writing or just describe the task.

On Windows, run the commands in Git Bash or WSL, or copy the folder into the same place with File Explorer.

Codex, Gemini CLI, Cursor and GitHub Copilot

These agents read the same SKILL.md folders. The one folder all four accept inside a project is .agents/skills/, so this copies the skill there:

git clone --depth 1 https://github.com/conorbronsdon/avoid-ai-writing.git
mkdir -p .agents/skills/avoid-ai-writing
git -C avoid-ai-writing archive HEAD | tar -x -C .agents/skills/avoid-ai-writing
AgentProject folderYour user folderNotes
OpenAI Codex
.agents/skills/
~/.agents/skills/
Type $ or run /skills to pick a skill; Codex also chooses one when your task matches.
Gemini CLI
.agents/skills/
.gemini/skills/
~/.agents/skills/
~/.gemini/skills/
Or install from Git with gemini skills install <git URL> --path <folder> --consent. Google replaced Gemini CLI with Antigravity CLI for free-tier and Google One users on 18 June 2026; Antigravity uses .agents/skills/ in a project and ~/.gemini/config/skills/ for you.
Cursor
.agents/skills/
.cursor/skills/
~/.agents/skills/
~/.cursor/skills/
Type / in Agent chat to pick a skill. Cursor also reads Claude Code's .claude/skills/ folders.
GitHub Copilot
.github/skills/
.agents/skills/
.claude/skills/
~/.copilot/skills/
~/.agents/skills/
Works in Copilot CLI, the cloud agent, and agent mode in VS Code, Visual Studio and JetBrains (IDE support depends on the version). Or run gh skill install OWNER/REPOSITORY SKILL (public preview, GitHub CLI 2.90.0 or later).
Only install skills you trust. Anthropic's advice is to use skills only from trusted sources, and to read every file of any other skill before using it: a skill can direct the agent to run commands, read files or send data in ways its description does not mention.

Sources for these steps

Checked against each vendor's documentation on 2026-11-14:

Preservation and editorial judgment

Mechanical checks can catch changed names, numbers and links; they cannot certify meaning. The agent must compare the final text with the source.

When a tool cannot run, the report must call that check unavailable or model-only rather than claiming verification.

Questions about avoid-ai-writing

Is it an AI authorship detector?

No. Human writing can trigger the same patterns; the source explicitly warns against consequential authorship verdicts.

Does it work equally well in every language?

No. Its rules and detector are calibrated for English. Other languages need their own evidence and rules.

Can detect mode change my draft?

No. Detect-only mode reports justified findings without rewriting or normalizing punctuation.

How many rewrites can it make?

At most two editing passes by default, or one when requested. Auditing unchanged text uses zero editing passes.

More AI skills

Related skills

Skill categories