Security audit
Security checks for vulnerabilities and agentic risk
This is a Markdown-only writing editor skill that can rewrite documents, with some style-scope caveats but no evidence of hidden execution, persistence, exfiltration, or privilege abuse.
Install only if you want an agent to actively rewrite text to sound less AI-generated. Review outputs before accepting them, especially for professional, academic, legal, factual, or ghostwritten material where preserving voice, neutrality, and authorship matters.
The phrase "Or invoke directly when editing documents" does not specify what exact trigger phrases, contexts, or constraints cause the skill to activate. Because it is broad and lacks negative examples or scope limits, it could lead to unintended invocation during ordinary editing activity.
The activation condition is broad enough to trigger on routine editing or review tasks, even when the user did not ask to disguise AI authorship or alter stylistic authenticity. In context, this can cause the agent to rewrite content more aggressively than intended, increasing the risk of unintended voice changes, provenance-masking, or policy-sensitive transformation of text.
The skill explicitly directs the agent to inject opinions, first-person perspective, humor, and 'personality' without requiring user opt-in. That can override the requested tone, distort the author's intent, and fabricate subjective stance where none existed, which is especially risky for professional, academic, legal, or factual content.
No suspicious patterns detected.
