Humanizer
v0.1.0Remove 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· 292·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name, description, and the SKILL.md all describe text editing to remove AI-style patterns; required binaries/env/configs are none. The declared capabilities (identify patterns, rewrite, preserve voice) are consistent with a text 'humanizer'.
Instruction Scope
The SKILL.md stays on-topic (scan input text for AI patterns and rewrite). However the allowed-tools list includes Read, Write, Edit, Grep, and Glob — tools that can access and search the filesystem. The instructions do not explicitly constrain processing to only user-provided text, so an agent with file access could be instructed (or misuse could occur) to scan or modify files beyond the intended input. Recommend limiting runtime file access or explicitly restricting the skill to operate only on supplied text.
Install Mechanism
Instruction-only skill with no install spec and no code files — nothing is downloaded or written to disk during install, which is the lowest-risk pattern.
Credentials
No environment variables, credentials, or config paths are requested, which is proportional to a text-editing skill.
Persistence & Privilege
always is false and the skill is user-invocable; it does not request persistent privileges or attempt to modify other skills or system settings.
Assessment
This skill appears coherent for editing text to remove AI-style phrasings and add a more human voice. Before installing or enabling it, consider: (1) the allowed-tools (Read/Write/Edit/Grep/Glob) let an agent access files — ensure the agent is only given the specific text you want edited and not broad filesystem permissions; (2) ethical and legal risks — this tool explicitly instructs removal of AI-origin signals, which can be misused to obscure provenance (avoid using on content where disclosure is required or on others' copyrighted/private material); (3) audit any agent actions that run the skill (confirm which files or inputs the skill will touch). If you want lower risk, permit the skill to operate only on pasted/text-provided inputs and deny general file-search or write privileges.Like a lobster shell, security has layers — review code before you run it.
latestvk97acw5vmqbybqcx8hvgb9prk981vpa7
License
MIT-0
Free to use, modify, and redistribute. No attribution required.