Security audit

Whatsapp Ultimate

Security checks for vulnerabilities and agentic risk

Overview

This WhatsApp skill describes powerful messaging, group-management, and chat-search actions, but its scope and safety boundaries are too unclear for automatic approval.

Review carefully before installing. Only use this skill where WhatsApp account access, recipients, groups, and chat-history searches are explicitly authorized, and require confirmation before sending messages or modifying groups.

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 (6)

Intent-Code Divergence

Medium

Confidence: 92% confidence
Finding: The skill documentation contradicts itself on whether an API key is required, which can cause operators to misconfigure authentication and accidentally rely on insecure defaults or unsafe workarounds. In a messaging skill that may send or search communications, ambiguity around auth requirements increases the chance of unauthorized use or failed security controls.

Intent-Code Divergence

Medium

Confidence: 95% confidence
Finding: The manifest mixes WhatsApp messaging functionality with unrelated SEO/ranking use cases, indicating the skill definition is confused, templated, or repurposed without proper review. This can trigger the skill in inappropriate contexts and lead an agent to perform communication actions when the user intended a different category of task.

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: Claiming SEO/ranking features in the core capability summary contradicts the rest of the skill and makes the operational scope unreliable. Scope confusion is dangerous for agent skills because it can cause incorrect tool selection and execution of message-sending or data-search behavior in contexts where those actions are unnecessary or privacy-invasive.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The activation guidance is overly broad, suggesting use for generic efficiency, automation, and workflow optimization tasks rather than tightly scoped communication actions. Overbroad routing increases the chance that an agent invokes a skill with exec/write/read abilities in unrelated situations, which expands the attack surface and risk of unintended actions.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: Conflicting 'Use when' guidance makes it unclear when the skill should be selected and combines communication and SEO contexts in the same manifest. This ambiguity is especially risky because the skill advertises command execution and messaging capabilities, so mistaken activation could expose data or send unintended outbound communications.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The skill documents message sending, group management, and history search without clear up-front warnings about privacy, consent, recipient impact, or authorization boundaries. Because these operations can access conversation history and send outbound content to individuals or groups, missing safeguards materially increase the risk of privacy violations, spam, or misuse of sensitive communications.

Static analysis

No suspicious patterns detected.