一个会逐渐成长的AI写作者

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:24 PM UTC
• Version:
  1.0.0

Overview

This is an instruction-only writing persona skill with no code or system access, though users should be transparent when publishing AI-assisted work.

This skill is safe to install as a code-free writing assistant. Use care with reader_feedback because comments can contain personal information, and disclose AI assistance where platform rules, professional norms, or reader expectations require it.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Low

• Confidence: 89% confidence
• Finding: The skill explicitly invites users to provide reader feedback and states it can use comments to evolve its writing style, but it provides no warning that such feedback may contain sensitive personal data. This creates a privacy risk because users may paste identifiable or confidential comments into the skill without any guidance on minimization, consent, or retention.

Natural-Language Policy Violations

Medium

• Confidence: 93% confidence
• Finding: The skill’s stated goal is to avoid 'AI-like' writing and make output feel like it was written by a human, without any mention of disclosure or user-controlled transparency. In context, this can facilitate deceptive content creation, impersonation-style misuse, or undisclosed synthetic authorship on social platforms and publishing channels.

VirusTotal

66/66 vendors flagged this skill as clean.