Security audit

WhatsApp消息工具(免费版)

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a disclosed WhatsApp command helper, but it also advertises unrelated SEO triggers that could route an agent into messaging or chat-search actions for the wrong task.

Install only if you explicitly want an agent to operate WhatsApp through wacli. Confirm recipients, message content, file paths, and chat-search requests before execution, protect ~/.wacli/ as sensitive account and message data, and remove the unrelated SEO invocation language before publishing or relying on automatic skill routing.

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 (5)

Description-Behavior Mismatch

Medium

Confidence: 96% confidence
Finding: The manifest description advertises SEO optimization, ranking, and search-traffic use cases even though the skill actually performs WhatsApp messaging and message search. This semantic mismatch can cause an agent to invoke a messaging tool in unrelated contexts, creating a confused-deputy risk where user data or outbound messages are handled under false pretenses.

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: The capability coverage statement claims support for SEO-related scenarios and broad keywords that are not actually implemented by the tool. In agent ecosystems that route skills based on keyword matching, this can incorrectly increase selection probability and lead to misuse of messaging, search, or local data access functions in irrelevant tasks.

Vague Triggers

High

Confidence: 98% confidence
Finding: The description contains an overly broad invocation condition telling agents to use this skill for SEO optimization, keyword analysis, ranking improvement, and traffic optimization. Because the skill has exec capability and can access/send WhatsApp content, broad misrouting materially raises the chance of unintended command execution, data exposure, or contacting third parties without appropriate user intent.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The capability coverage keywords define a very wide trigger surface with fragmented and loosely related terms, including SEO and generic 'Use/when' phrasing. This broadens automatic matching beyond the tool's real function and increases the likelihood that an agent activates a powerful exec-backed messaging skill in the wrong context.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill describes message-history search and local credential/data storage but does not foreground a clear privacy warning near those capabilities. Users or agents may not appreciate that chat contents, media, and authentication material are stored locally and searchable, which raises confidentiality and consent risks if the tool is invoked casually.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.