ClawHub
Back to skill

Security audit

Java 工具箱专业版

Security checks across malware telemetry and agentic risk

Overview

This Java toolkit does not show malicious behavior, but it is routed far too broadly for a skill that can run commands and write files.

Install only if you intend to use it for Java repository scanning, ruleset management, JVM/GC review, and build or coverage gates. Treat command execution and file writes as requiring explicit user direction, especially if the current task is generic reporting, planning, or non-Java data analysis.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The trigger condition is materially mismatched to the skill's stated Java scanning and JVM tuning purpose, encouraging invocation for generic data analysis and reporting tasks outside its intended scope. In an agent environment with read/exec/write tools, overbroad routing can cause the skill to run in inappropriate contexts and perform filesystem or command actions on unrelated tasks.

Vague Triggers

High
Confidence
96% confidence
Finding
An overly broad trigger increases the chance that the agent selects this skill for requests far outside Java analysis, despite the skill advertising exec/write capability and command-driven workflows. Mis-selection is dangerous because it can lead to unintended command execution, file modification, or misleading outputs in contexts where the user did not ask for repository scanning or build actions.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The manifest description includes vague guidance such as project management, task planning, and team collaboration, which overlaps with common non-code requests and can broaden when the skill is auto-invoked. In combination with powerful tools, this scope creep can cause an agent to apply code-scanning behavior or command execution in situations unrelated to Java engineering tasks.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.