Back to skill
Skillv1.0.0
ClawScan security
AI Text Humanizer — Remove AI Writing Patterns · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 6, 2026, 6:39 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill is an instruction-only text-editing helper whose requirements and runtime instructions match its stated purpose and ask for no extra privileges or installs.
- Guidance
- This skill is internally consistent and low-risk: it only contains editing instructions and doesn't request credentials or install code. Before using it, consider: (1) review its rewrites for subtle meaning changes—the 'Add soul' guidance may introduce first-person opinion or rhetorical flourishes; (2) avoid pasting sensitive or confidential text into any third-party skill or tool; (3) if you want tighter control, keep it user-invocable only (do not allow unattended/autonomous runs) so you can approve changes before they are applied; and (4) the README cites Wikipedia sources and the skill is instruction-only (MIT license in README), but test on non-sensitive samples to confirm the editing style matches your expectations.
Review Dimensions
- Purpose & Capability
- okName, description, README, and SKILL.md all consistently describe a text 'humanizer' that finds AI-style patterns and rewrites text. There are no unrelated env vars, binaries, or config paths requested that would be disproportionate to an editing task.
- Instruction Scope
- okSKILL.md explicitly instructs the agent to scan provided text for listed AI patterns and rewrite passages to preserve meaning and voice. Allowed-tools (Read, Write, Edit, Grep, Glob, AskUserQuestion) are appropriate for an editing skill. The instructions do not direct the agent to read arbitrary system files, access credentials, or send data to external endpoints.
- Install Mechanism
- okNo install spec and no code files (instruction-only). Nothing will be downloaded or written to disk by an installer—this is the lowest-risk model.
- Credentials
- okThe skill requests no environment variables, credentials, or config paths. That is proportionate to a local text-editing instruction-only skill.
- Persistence & Privilege
- notealways:false (default) and user-invocable:true. disable-model-invocation:false (agent may invoke autonomously, which is the platform default). This is expected for a helper skill, but note that autonomous invocation means the agent could call the humanizer whenever it deems relevant.