Humanize AI text

MaliciousAudited by ClawScan on May 10, 2026.

Overview

This skill is explicitly designed to disguise AI-generated writing so it can evade AI-detection systems, although its code otherwise appears to run locally on user-selected files.

Avoid this skill if your goal is honest editing or compliance with disclosure rules. Its main purpose is to make AI-generated text harder to detect, which can deceive reviewers and violate policies, even though the provided scripts appear to operate locally without network or credential access.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Using this skill could help misrepresent AI-generated writing as human-written and violate academic, publishing, workplace, or platform policies.

Why it was flagged

The stated purpose is not merely style improvement; it is to make AI-generated text evade detection systems, which can mislead reviewers, educators, publishers, or other relying parties.

Skill content
description: Humanize AI-generated text to bypass detection... pass AI detectors like GPTZero, Turnitin, and Originality.ai... Makes robotic AI writing undetectable and human-like.
Recommendation

Do not install or use this skill for bypassing detection. Use transparent editing or proofreading tools and disclose AI assistance where required.

What this means

A user or agent could modify local writing files if directed to run the script with output paths.

Why it was flagged

The transformer reads local input text and can write transformed output files. This is expected for the stated purpose, but users should understand it can create or overwrite files at paths they provide.

Skill content
text = Path(args.input).read_text() if args.input else sys.stdin.read() ... Path(args.output).write_text(result)
Recommendation

Run it only on files you intend to process, write to a new output path, and review changes before replacing originals.

What this means

Users have limited external provenance information for deciding whether to trust the package.

Why it was flagged

The skill has no disclosed source homepage or install provenance. The included code is small and self-contained, so this is a provenance note rather than evidence of hidden installation behavior.

Skill content
Source: unknown; Homepage: none; Install specifications: No install spec
Recommendation

Review the included scripts before use and prefer skills with clear source provenance and declared runtime requirements.