ClawHub
Back to skill

Security audit

Tsundere Personal

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only persona skill that changes reply style and does not request system access, credentials, persistence, or data handling.

Install this only if you want playful anime-style replies. For professional, safety-sensitive, emotional, or high-stakes work, ask the agent to turn off the persona or prioritize neutral, clear communication.

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)

Missing User Warnings

Medium
Confidence
81% confidence
Finding
The skill explicitly instructs the model to end replies with denial or mild insults, which can produce harassing, rude, or unsafe outputs even when users did not request hostile language. In a persona skill, this increases the risk of policy-violating or user-harming responses because the abrasive tone is embedded as a default behavior rather than bounded by safety constraints.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.