Security audit

mar-document-pro

Security checks across malware telemetry and agentic risk

Overview

This is a coherent document-processing skill, but users should know that AI analysis sends extracted document text to the SkillBoss API.

Install only if you are comfortable having document text analyzed through SkillBoss API Hub. Avoid using it on confidential, regulated, or third-party documents unless you have permission to send their contents to that provider, and prefer a dedicated API key with limited access.

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 (3)

Vague Triggers

Medium

Confidence: 86% confidence
Finding: 技能元数据中的描述将“分析文档、提取内容、总结报告”等常见需求直接作为触发条件，但没有提供足够的限定条件或确认步骤，容易在普通对话中被误触发。误触发后该技能可能读取用户提供的文档并进一步调用外部 API，放大了隐私暴露和越权处理的风险。

Vague Triggers

Medium

Confidence: 88% confidence
Finding: 触发场景列出了“分析”“总结”“提取内容”“询问具体信息”等宽泛表述，缺少文件已提供、格式受支持、是否允许外发等边界约束。这会使技能在高频自然语言场景中被错误激活，并可能对敏感文档执行读取、解析和外部传输。

Missing User Warnings

High

Confidence: 98% confidence
Finding: 技能明确说明会在提取文档文本后将内容发送到 SkillBoss API Hub 做摘要分析，但没有提供清晰的数据出境、第三方处理、保留策略或用户同意说明。由于文档经常包含合同、财务、身份信息或内部资料，这种默认外发会带来明显的隐私、合规和数据泄露风险。

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.