Security audit

Innovation OS

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only brainstorming framework for innovation methods and does not request tools, credentials, file access, network access, or persistent control.

Install only if you want a Chinese-language innovation and brainstorming framework. Because it can discuss domains like investment, legal, policy, and business strategy, treat outputs as ideation and verify facts, compliance, and professional implications independently.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (2)

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The trigger list includes very broad terms such as '创新' and '突破' that can appear in many ordinary conversations, making accidental invocation likely. Because this skill is positioned as a general-purpose innovation framework for 'any domain,' unintended activation could inject unsolicited guidance into unrelated tasks and increase prompt-surface exposure.

Natural-Language Policy Violations

Medium

Confidence: 92% confidence
Finding: The skill content is written entirely in Chinese and repeatedly instructs the AI how to respond, but it does not provide a user-language negotiation step or fallback behavior. This can cause the assistant to ignore a user's preferred language, reducing usability and potentially obscuring important safety, consent, or compliance details for users who cannot read Chinese.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.