ClawHub

Humanizer 1

v1.0.0

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comp...

⭐ 0· 76·3 current·3 all-time
by三昧@hanxueda·duplicate of @lion504/humanizer-1-0-0
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description (remove AI-generated writing patterns and add voice) match the SKILL.md guidance, examples, and transformation steps. There are no unrelated env vars, binaries, or install steps requested.
ℹ
Instruction Scope
SKILL.md gives explicit, narrow instructions about detecting patterns and rewriting text; it does not instruct the agent to read system configs or environment variables. The skill's allowed-tools list includes file and grep/glob operations (Read, Write, Edit, Grep, Glob), which could be used to access files if the agent is granted filesystem access — but the skill itself does not tell the agent to scan arbitrary system files.
✓
Install Mechanism
No install spec and no code files — instruction-only skill. This minimizes disk write and supply-chain risk.
✓
Credentials
Requires no environment variables, credentials, or config paths. Requested access is proportionate to an editing tool.
✓
Persistence & Privilege
always is false and the skill is user-invocable. The skill does not request permanent presence or modification of other skills or system settings.
Assessment
This skill appears internally consistent and safe in that it only provides editing guidance and requests no credentials. Before using it, avoid feeding sensitive secrets or private files for automatic humanization; if you allow the agent filesystem access, consider restricting which documents it may read. Review outputs before publishing (the skill encourages adding opinions/first-person voice, which may unintentionally change factual tone), and enable user confirmation when invoking if you don't want autonomous edits applied without review.

Like a lobster shell, security has layers — review code before you run it.

latestvk97dgxwb12ypyyr2t4hzpwavps83jvfs

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments