Avoid AI Writing - Setup Guide¶

Source: conorbronsdon/avoid-ai-writing GitHub: conorbronsdon/avoid-ai-writing Skills: 1 skill · 1.6K installs Category: Content Quality First Seen: catalogued August 17, 2026 sweep (avoid-ai-writing on skills.sh since March 7, 2026) Quality Tier: 🟡 Trusted - Snyk Warn named (see Limitations)

Avoid AI Writing audits content for AI writing patterns ("AI-isms") and rewrites them out. Its distinguishing trait is epistemic honesty: the skill opens by stating it is a writing-quality tool, not a verdict - citing the Stanford false-positive study on non-native English writers and the 2025 open-source detector misclassification research. Patterns are signals to pair with context, not proof.

Installation¶

npx skills add conorbronsdon/avoid-ai-writing

Individual skills install with the explicit repo form:

npx skills add https://github.com/conorbronsdon/avoid-ai-writing --skill avoid-ai-writing

Prerequisites¶

Requirement	Details
Node.js + npx	For the skill installer
Prose to audit	Any draft - post, email, doc, or page copy

What It Provides¶

Three operating modes:

Mode	Purpose
`rewrite` (default)	Flag AI-isms and rewrite the text to fix them, including a built-in corrective second pass
`detect`	Flag AI-isms only - no rewriting ("scan," "flag only," "audit only")
`edit`	Edit a file in place with minimal, targeted changes; preserves already-human passages

Optional voice profiles (casual / professional / technical / warm / blunt), context hints (linkedin, blog, investor-email, docs), and an iterate-to-convergence pass capped at 2 rounds round out the CLI surface.

The skill's built-in guardrails warn against using its signals as the sole basis for consequential decisions (academic integrity, hiring, publication) and note that the same patterns fire on second-language writing, deadline-pressed humans, and compressed technical genres.

Quick Start¶

Install: npx skills add conorbronsdon/avoid-ai-writing
Ask: "audit this draft with avoid-ai-writing and rewrite it"
Review the flagged patterns; accept or reject each rewrite

CorpusIQ Use Cases¶

Use Case	How
Already installed in production	This skill runs in the CorpusIQ Hermes profile - this guide closes the docs-catalog gap
Public content gate	Pairs with content-voice rules and slop scoring before anything ships publicly
Outreach email quality	AI-ism audit on cold outreach and partner emails keeps the human voice

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

Limitations / Verification¶

Security audits on avoid-ai-writing: Gen Agent Trust Hub Pass, Socket Pass, Snyk Warn - named per catalog policy
Publisher-page install counts verified (1.6K); GitHub 3.0K stars on the repo
Install count is below the usual drafting bar - catalogued because the skill is installed in the CorpusIQ production profile and was flagged as a docs gap by the August 17 ecosystem scan
Signals are statistical, not proof - the skill itself says so and cites the research

npx skills add conorbronsdon/avoid-ai-writing   # verify install works

Powered by CorpusIQ