Humanizer by JZ
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This is an instruction-only writing-editing skill with no code, credentials, or install scripts; the main cautions are that it can edit files and is explicitly meant to make AI-styled text appear more human.
This appears operationally low risk because it is instruction-only and requests no credentials or install-time code. Before installing, consider whether you are comfortable granting it document read/edit capability, review any file changes it makes, and avoid using it to misrepresent authorship where disclosure matters.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
64/64 vendors flagged this skill as clean.
Risk analysis
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.
If used on files rather than pasted text, the agent may change drafts or documents on disk.
These tools allow the agent to find, read, and modify local files. That is expected for an editing skill, but users should ensure edits are limited to intended documents.
allowed-tools: - Read - Write - Edit - Grep - Glob - AskUserQuestion
Use the skill on specific text or clearly identified files, and review diffs or backups before accepting changes.
Outputs could make AI-assisted text seem more personally or human-authored than it is.
The skill is openly designed to make AI-styled writing appear more human-written, which can affect reader trust or authorship expectations.
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written.
Use it for legitimate editing, but follow any disclosure, academic, workplace, or publication rules about AI-assisted writing.