Security audit

Whatsapp Style Guide Free

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a static WhatsApp formatting guide, but it asks for command execution despite having no bundled scripts and contains unrelated/generated sections that make its real scope unclear.

Review before installing. This does not show evidence of theft, persistence, or destructive behavior, but users should treat it as an over-scoped static guide: avoid letting it run shell commands, and rely on it only as a Chinese WhatsApp formatting reference after checking examples for accuracy.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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 (4)

Intent-Code Divergence

High

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill presents itself as a static WhatsApp formatting guide, but the later Bluetooth/product-parameter sections introduce unrelated capabilities and pseudo-operations. This inconsistency is dangerous because it expands the apparent scope of the skill beyond its declared purpose, making it harder for users and agents to reason about what actions are appropriate and increasing the chance of misuse or prompt confusion.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: Examples labeled as WhatsApp-compliant include syntax such as Markdown headings that the guide itself marks as unsupported. This is dangerous because it teaches contradictory behavior, causing agents or users to generate malformed messages and undermining trust in the skill's ruleset.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The skill requests exec-capable command-line usage even though its stated function is a static reference guide. Unnecessary execution capability increases attack surface because an agent may be induced to run local commands for a task that should require only passive documentation access.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The configuration example sets the guide language to zh_CN, and the skill content is written entirely in Chinese without any stated option to choose another language in the free version. This creates a natural-language locale constraint that is not presented as user-selectable or clearly justified as a region-specific tool.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.