ClawHub
Back to skill

Security audit

阅读教学指导

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Chinese reading-teaching guide with templates and no executable code; users should just be mindful of student record updates and broad trigger phrases.

Install only if you want an agent to help with Chinese reading lesson design. Use pseudonyms for students, avoid entering unnecessary personal information, and confirm before allowing any connected student-analysis or resource-library records to be updated.

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
93% confidence
Finding
The activation examples are very broad, common classroom phrases such as asking how to teach a text or reading comprehension. In an agent ecosystem, this can cause the skill to trigger in contexts where the user did not explicitly request it, leading to over-collection of context, inappropriate workflow injection, or interference with other skills.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger timing section defines activation through ambiguous scenarios like '学生读不懂怎么办' and '怎么教思辨', without clear boundaries for when this skill should defer. That increases the chance of unintended activation and cross-skill collisions, especially because the skill also writes data to downstream systems such as student-analyzer and resource-library.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.