ClawHub

渣男

Security checks across malware telemetry and agentic risk

Overview

This text-only skill helps analyze manipulative relationship language and contains dual-use examples, but it is clearly framed for recognition and defense rather than hidden or harmful execution.

Install only if you want a Chinese-language framework for spotting manipulative relationship communication. Treat the example tactics as red flags to recognize, not scripts to use, and avoid relying on the skill alone for serious relationship, abuse, legal, or mental-health decisions.

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)

Ssd 4

Medium
Confidence
89% confidence
Finding
This section encodes a compact decision tree for deception, minimization, blame-shifting, and emotional coercion. Even though the skill is framed as recognition/analysis, the explicit if-then playbook materially lowers the barrier for reproducing manipulative behavior, especially because it is presented as reusable heuristics rather than solely as warning indicators.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal