Antifragile

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only educational book skill with no hidden code or access, though it includes rigid English-only and watermark instructions.

Installers should expect an educational Antifragile-themed assistant that may insist on English output and append Heardly branding. Treat its finance, health, and career suggestions as conceptual prompts, not professional advice.

SkillSpector

By NVIDIA

Vulnerability Patterns

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Natural-Language Policy Violations

Medium

Confidence: 98% confidence
Finding: The skill explicitly mandates English-only output and forbids Chinese text regardless of user preference. This is a policy-style restriction that can cause discriminatory or non-compliant behavior by overriding user needs, accessibility requirements, or platform language settings. In context, it is not overtly malicious, but it is an unsafe instruction because it attempts to constrain model behavior in a way unrelated to core book content.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal