ClawHub

LLM 发展趋势报告 (Ray1012)

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only LLM trend-report skill with minor usability caveats but no code, credential access, persistence, or hidden high-impact behavior.

Installers should treat this as a low-risk reporting aid. It may activate for broad AI-frontier or technology-prediction prompts, appears to default to Chinese output, and asks the agent to search current public sources, so verify important claims and be aware that one optional referenced paper list is missing from the package.

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
90% confidence
Finding
The activation text uses very broad triggers such as questions about 'AI frontier', 'technology predictions', and 'LLM development direction', which can match many ordinary requests and cause the skill to activate outside its intended scope. Over-broad activation increases the chance of prompt hijacking, unexpected tool use, or the skill overriding more appropriate system behavior in unrelated conversations.

Natural-Language Policy Violations

Medium
Confidence
82% confidence
Finding
The skill description and content are written to produce Chinese-language output by default without offering a user-language choice or documenting a strict locale requirement. While not directly a code-execution issue, this can cause misleading or inaccessible responses, reduce user awareness of what the skill is doing, and create unsafe UX in multilingual contexts where precision and consent matter.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal