ClawHub
Back to skill

Security audit

设计清晰一致视觉精致的用户界面

Security checks for vulnerabilities and agentic risk

Overview

This UI design skill is mostly a design guide, but it asks for broad read, write, and command-execution authority that is not clearly scoped to UI work.

Review this skill before installing if your agent grants tool access. Allow it only in projects where reading and writing UI files and running explicit build/test commands is acceptable, and avoid providing broad API keys or access to unrelated private files.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill is presented as a UI design toolkit, but its documentation advertises broad file I/O, external API access, and especially system command execution capabilities that are not tightly scoped to that purpose. In an agent environment, this expands the attack surface and can let design-oriented prompts pivot into arbitrary local actions, data access, or outbound requests if the host grants those tools.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

Claiming command execution happens in a 'safe sandbox' without any accompanying isolation mechanism, policy, or configuration can mislead users and reviewers into overtrusting dangerous behavior. If no real sandbox exists, users may approve execution they otherwise would reject, leading to unintended command execution on the host environment.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
80% confidence
Finding

The invocation language is overly broad and says to use the skill for general design creation, UI design, poster making, brand visuals, and automation workflows without clear activation boundaries. In agent systems, vague triggers increase the chance the skill is invoked in contexts beyond its intended scope, which is more dangerous here because the skill also advertises read/write/exec capabilities.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.