ClawHub
Back to skill

Security audit

批处理专家

Security checks across malware telemetry and agentic risk

Overview

This skill provides disclosed batch-processing guidance with safeguards and no hidden executable payloads.

Install this if you want an agent to help design safer batch jobs. Before using it on deletion, messaging, payment, account, or production-data workflows, require a small dry run, review the exact command or script, confirm checkpoint location, and verify rollback or retry behavior.

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 (1)

Vague Triggers

Medium
Confidence
96% confidence
Finding
The trigger keywords include very generic terms such as '批量', '并行', '流式', '进度', and 'OOM', which are common across many unrelated tasks. In an agent ecosystem, this can cause unintended activation of the skill in contexts where batch execution or command generation is not appropriate, increasing the chance that the agent suggests or performs broad `exec`-enabled operations without explicit user intent.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.