Security audit

duange-zero-skill

Security checks across malware telemetry and agentic risk

Overview

This skill helps Chinese-speaking beginners create Codex skills; its broad trigger phrases could cause occasional misrouting, but the package is transparent and contains no executable or hidden data-access behavior.

Reasonable to install for drafting Codex Skills, especially for Chinese-speaking beginners. Before using generated skills broadly, review and tighten their trigger phrases, keep explicit non-trigger examples, and preserve confirmation rules for deletion, overwriting, posting, payments, authorization, private files, and credentials.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger phrases are very generic and map to common user requests such as wanting to make a helper or asking for step-by-step assistance. This can cause the skill to activate in contexts where the user did not specifically intend to invoke this skill, leading to misrouting, confused behavior, or accidental exposure of skill-specific capabilities.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The description includes broad trigger phrases like '想做 Skill' and '创建小助手', which can overlap with ordinary conversation and cause accidental activation. Unintended invocation can route unrelated user input into this skill, producing confusing behavior and potentially causing the skill to solicit sensitive workflow details or generate files when the user did not intend to use it.

Vague Triggers

Medium

Confidence: 82% confidence
Finding: The suggested invocation examples are generic productivity requests without enough bounding context, so they may match many ordinary tasks that should be handled by other skills or normal chat. This increases the chance of over-triggering and skill confusion, which is especially relevant because this skill is a meta-creator that can steer users into generating project files and workflows they did not explicitly request.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The example trigger phrases are broad, conversational, and closely overlap with normal user language about creating assistants or skills. In this skill context, that increases the chance of accidental invocation, causing the agent to enter a skill-construction workflow when the user may have only been casually discussing ideas rather than intentionally activating the skill.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The prompt asking users how they will 'call it out later' collects a trigger phrase but does not constrain format, uniqueness, or safe matching rules. That can lead users to choose vague everyday language as an activation condition, making unintended triggering more likely across unrelated conversations.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.