ClawHub
Back to skill

Security audit

HTML编码工具-免费版

Security checks for vulnerabilities and agentic risk

Overview

This is mainly a Markdown-only HTML coding helper, but its routing and credential notes are sloppy and should be treated cautiously.

Install only if you want an HTML-focused helper, and keep its use limited to HTML markup, forms, accessibility, responsive images, and explicit validation tasks. Do not provide API keys unless the host agent itself requires them, and review any proposed shell command before allowing it to run.

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)

Description-Behavior Mismatch

Medium
Confidence
90% confidence
Finding
The manifest advertises a narrowly scoped HTML coding skill, but the body expands usage to broad development, deployment, automation, batch processing, and command execution scenarios. This scope drift can cause an agent to invoke the skill in contexts far beyond HTML generation, increasing the chance of unsafe tool use and user confusion about what the skill is authorized to do.

Intent-Code Divergence

High
Confidence
96% confidence
Finding
The documentation gives conflicting security-critical setup guidance: one section says no extra API key is required, while another says the skill cannot be used without a key. Contradictory instructions about secrets handling can lead operators to improvise insecure workarounds, expose credentials, or misconfigure the skill in production environments.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger conditions are broad enough to match common development, debugging, testing, and deployment requests, despite the skill claiming to be an HTML-focused assistant. Over-broad routing increases unintended invocation, which matters more here because the skill declares exec/browser/read tools and discusses command execution, potentially widening the operational attack surface.

Static analysis

No suspicious patterns detected.