ClawHub
Back to skill

Security audit

插件UI架构

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to be a loosely scoped plugin UI helper, but it asks for broad write, command, API, and credential-related authority without enough concrete boundaries.

Review carefully before installing. Only use this skill in a disposable or controlled SkillHub environment, and do not provide API keys or allow shell/write actions unless you have separate, trusted installation instructions and know exactly which files will be changed.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
91% confidence
Finding
The skill is presented as a UI plugin architecture helper, but later advertises broad powers like file processing, external API integration, and command execution. That mismatch increases the chance an agent or user will grant it broader operational authority than necessary, enabling unintended file changes, outbound requests, or shell actions under an overly generic skill description.

Vague Triggers

Medium
Confidence
84% confidence
Finding
An overly broad and ambiguous invocation description can cause accidental triggering in unrelated contexts. In a skill that declares read/write/exec tools, vague activation language is dangerous because it may cause an agent to perform impactful actions without the user intentionally selecting this specific skill.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The example trigger phrase is so generic that many unrelated tasks could match it. Because the skill advertises operational capabilities beyond simple advice, ambiguous triggering raises the risk of unintentional execution, especially in automated skill-selection environments.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The document advertises file writing, API calls, and command execution without clear upfront warnings, approvals, or boundaries. In an agent environment, this can normalize high-impact actions and lead users or orchestration systems to invoke the skill without understanding that it may modify files, contact external services, or run shell commands.

Static analysis

No suspicious patterns detected.