ClawHub

Learn Prototype

Security checks across malware telemetry and agentic risk

Overview

This is a Chinese-language coaching skill for iterative learning and prototyping, with no code execution or sensitive access requested.

Install this if you want a Chinese-language coaching workflow for learning by prototyping and iterative improvement. Be aware it may trigger on many general learning or optimization requests, so users who prefer another language or a narrower workflow may want clearer activation rules.

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 (2)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger conditions are very broad ('要做 X、研究 X、提升 X、怎么优化 X') and can match a large fraction of ordinary learning, planning, and improvement requests. In an agent environment, this can cause the skill to activate when more appropriate skills or default behaviors should handle the task, leading to misrouting, reduced reliability, and unexpected behavior across many user interactions.

Natural-Language Policy Violations

Medium
Confidence
82% confidence
Finding
The skill content is written entirely in Chinese and frames the interaction in Chinese without indicating that the agent should adapt to the user's preferred language. If selected for users operating in another language, it can degrade usability, cause misunderstanding of instructions, and create inconsistent agent behavior that is difficult for the user to correct.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal