Skill Detail

Audit and rewrite AI-sounding prose with Avoid AI Writing

Use Avoid AI Writing when an agent needs a repeatable audit, rewrite, or in-place edit workflow for removing AI writing patterns from draft prose.

Content Writing & SEOMulti-Framework
Content Writing & SEO Multi-Framework Security Reviewed
⭐ 2.6k GitHub stars ⬇ 23/wk npm
INSTALL WITH ANY AGENT
npx skills add agentskillexchange/skills --skill audit-and-rewrite-ai-sounding-prose-with-avoid-ai-writing Copy
Works best when you want a reusable capability, not another fragile one-off prompt.
At a glance
Tools required
Agent with SKILL.md support such as Claude Code, OpenClaw, Codex, Cursor, Hermes, or compatible agent runtime
Install & setup
For Claude Code, clone the repo into ~/.claude/skills/avoid-ai-writing or install the plugin marketplace from conorbronsdon/avoid-ai-writing. For OpenClaw, run clawhub install avoid-ai-writing or clone into ~/.openclaw/skills/avoid-ai-writing. For Codex, place SKILL.md under .agents/skills/avoid-ai-writing or ~/.agents/skills/avoid-ai-writing.
Author
Conor Bronsdon
Publisher
Independent
Last updated
Jul 23, 2026
Quick brief

Avoid AI Writing is a portable SKILL.md workflow for auditing and rewriting prose that sounds AI-generated. It gives agents three concrete modes: rewrite text, detect issues without rewriting, or edit a file in place with a verification report. Use it when the operator wants an agent to clean a draft, scan copy for AI tells, or preserve a house voice while removing boilerplate phrasing.

How it works

What this skill actually does

Invoke this instead of using a normal writing product when the work needs to happen inside an agent session, close to the draft file or prompt context, with a structured issue list, rewrite, change summary, and second-pass audit. The boundary is narrow: this entry is for the repeatable editing skill and its detection/edit modes, not for a generic content platform, broad writing advice, or a standalone detector listing.

Inputs and prerequisites: Agent with SKILL.md support such as Claude Code, OpenClaw, Codex, Cursor, Hermes, or compatible agent runtime.

Setup notes: For Claude Code, clone the repo into ~/.claude/skills/avoid-ai-writing or install the plugin marketplace from conorbronsdon/avoid-ai-writing. For OpenClaw, run clawhub install avoid-ai-writing or clone into ~/.openclaw/skills/avoid-ai-writing. For Codex, place SKILL.md under .agents/skills/avoid-ai-writing or ~/.agents/skills/avoid-ai-writing.

Source and verification boundary: use https://www.chainofthought.show/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.

Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.