ClawHub
Back to skill

Security audit

文件

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to be a file organizer, but its instructions are underspecified and include destructive duplicate deletion plus unrelated API/network guidance.

Review before installing. Use it only on non-critical folders or backups, require a preview before any moves or deletions, and do not provide API keys unless the publisher documents a specific required service and data flow.

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 (4)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The skill is presented as a local file organizer, but it also introduces unrelated API-key setup, LLM API dependency, and network-oriented guidance. This broadens the apparent trust boundary and can mislead an agent or user into granting unnecessary secrets or capabilities that are unrelated to file sorting, increasing the chance of over-privileged execution or secret exposure.

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
Advertising command execution, API integration, and network-security controls inside a file-organizer skill creates a capability mismatch that can normalize dangerous behavior and mask abuse behind an innocuous label. In an agent ecosystem, this makes the skill more dangerous because users may approve it expecting simple local file operations while it implies broader execution and communication powers.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The invocation guidance is overly vague and does not define when the skill should or should not be used, what inputs are allowed, or what safety constraints apply. In an automated agent setting, ambiguity can cause the skill to be selected in inappropriate contexts and operate on unintended files or tasks, especially given its write and exec tool access.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill advertises optional duplicate-file deletion without a strong warning, preview mode, or confirmation requirement. Because deletion is destructive and duplicate detection can be error-prone, this can lead to irreversible data loss if the agent applies the option automatically or a user misunderstands the behavior.

Static analysis

No suspicious patterns detected.