Humanize
v1.0.0Remove AI writing patterns from text. Use when editing, reviewing, or rewriting text to sound more natural and human-written. Detects patterns like inflated symbolism, promotional language, em dash overuse, AI vocabulary, and sycophantic tone.
⭐ 9· 5.4k·73 current·74 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description (remove AI writing patterns / humanize text) align with the provided SKILL.md. The skill is instruction-only and does not ask for binaries, credentials, or config paths that would be unrelated to rewriting text, which is proportionate for this purpose.
Instruction Scope
Runtime instructions are limited to scanning and rewriting the provided text according to listed patterns, preserving meaning and tone. The SKILL.md does not instruct the agent to read files, environment variables, or contact external endpoints. Note: the guidance encourages adding specificity and citations; that can lead to fabricated or invented sources if the agent isn't constrained to only use verifiable information—this is a content-integrity risk rather than a coherence/security issue.
Install Mechanism
No install spec and no code files are present, so nothing is written to disk and no external packages are fetched — lowest-risk delivery model.
Credentials
The skill requests no environment variables, credentials, or config paths. There are no unexpected permissions or secret requests relative to the stated functionality.
Persistence & Privilege
always:false (default). disable-model-invocation:false is the platform default and not by itself a problem. The skill does not request persistent system presence or modify other skills/configuration.
Assessment
This skill appears coherent and low-risk: it only contains instructions for rewriting text and asks for no credentials or installs. Before using: (1) do not submit secrets or sensitive personal data for rewriting, since the skill will process whatever you send; (2) review outputs for factual accuracy—the guidance to 'cite specific sources' or 'be specific' can cause the model to invent details or attributions if not otherwise constrained; (3) if you don't want the agent to invoke this skill autonomously, keep autonomous invocation disabled at the agent level; and (4) verify any added first-person opinions or claims before publishing, since the skill encourages adding voice which can unintentionally misattribute statements.Like a lobster shell, security has layers — review code before you run it.
latestvk973t9b2mt05s8djk67khpyv5180a50r
License
MIT-0
Free to use, modify, and redistribute. No attribution required.