ClawHub
Back to skill

Security audit

WhatsApp Styler

Security checks for vulnerabilities and agentic risk

Overview

This is a simple WhatsApp text-formatting guide with some sloppy metadata, but it does not contain hidden commands, persistence, credential handling, or data exfiltration behavior.

Install only if you want an agent to format messages specifically for WhatsApp. Be aware that the skill metadata uses overly broad trigger words and includes unnecessary execution-related boilerplate, so maintainers should narrow the triggers and remove the MD+execute claim.

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 Markdown-only WhatsApp formatting guide, yet it also claims 'MD+execute' capability and references command-line execution. This mismatch can cause an agent or operator to over-trust the skill with execution privileges it does not need, increasing the attack surface for accidental or unauthorized command execution.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger keywords include very broad everyday terms such as 'ensure', 'sent', 'messages', and 'skill', making accidental activation likely. Overbroad activation can cause the agent to apply this skill in unrelated contexts, creating prompt-routing confusion and opening the door for unintended instruction interference with other tasks.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The repeated trigger section restates broad keywords without defining activation boundaries, so the skill may trigger on common language absent any true WhatsApp-formatting request. In an agent ecosystem, ambiguous routing can misapply this skill, override more appropriate skills, or alter outputs unexpectedly.

Static analysis

No suspicious patterns detected.