ClawHub
Back to skill

Security audit

web-artifacts-builde

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to be a web artifact builder, but its instructions are broad and unclear enough that a user should review it before installing.

Install only if you are comfortable with a skill that can guide an agent using local read and command execution tools. Keep its use limited to explicit web artifact building tasks, review commands before execution, and avoid giving it unnecessary credentials or sensitive files.

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
88% confidence
Finding
The skill is presented as a web/HTML artifact builder, but the documentation also advertises broad API access, command execution, and file handling behaviors that are not tightly scoped to that purpose. This mismatch can mislead an agent into granting or using powerful capabilities under a benign-looking label, increasing the risk of overbroad execution and unsafe automation.

Vague Triggers

High
Confidence
95% confidence
Finding
The trigger condition states the skill should activate whenever a user needs to perform related operations, which is overly broad and ambiguous. In an agent environment with exec and read tools, such vague routing can cause the skill to run in unintended contexts and perform sensitive actions without sufficiently specific user intent.

Natural-Language Policy Violations

Medium
Confidence
80% confidence
Finding
The description says the skill supports Chinese interaction and is framed as using Chinese by default without any explicit user opt-in. Forcing or assuming a language can cause users or downstream agents to misunderstand prompts, confirmations, or safety-relevant outputs, which is especially risky when the skill may also drive command execution or file operations.

Static analysis

No suspicious patterns detected.