Security audit

DNS网络管理工具

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a DNS/networking helper, but it broadens into generic command, file, API, and credential handling without clear boundaries.

Review this skill before installing. Use it only for DNS, certificate, and network checks against systems you are authorized to test. Do not provide API keys or credentials unless the publisher documents the exact service and reason, and avoid letting it run generic file, API, or shell tasks outside the stated DNS/networking purpose.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (7)

Description-Behavior Mismatch

Medium

Confidence: 95% confidence
Finding: The skill is declared as a DNS/network debugging tool, but these sections expand its scope to generic file processing, API integration, and command execution. That mismatch can cause an agent or user to authorize broader capabilities than necessary, increasing the chance of unsafe execution or data handling outside the expected DNS troubleshooting context.

Description-Behavior Mismatch

Medium

Confidence: 94% confidence
Finding: The scenario and metrics content describes unrelated data parsing and aggregation workflows, which materially broadens the apparent use cases of the skill. In an agent setting, this can mislead orchestration logic into invoking the skill for tasks involving arbitrary content handling rather than limited network diagnostics.

Context-Inappropriate Capability

Medium

Confidence: 93% confidence
Finding: Introducing API key configuration and external API usage without clear necessity for DNS/network diagnostics creates an unjustified credential-handling surface. This can lead users or agents to expose secrets or permit outbound integrations that are not required for the stated task.

Intent-Code Divergence

Low

Confidence: 88% confidence
Finding: The request schema uses generic content-processing fields instead of DNS/network-specific inputs such as hostnames, ports, resolvers, or protocols. This ambiguity weakens input boundaries and makes it easier for the skill to be repurposed for arbitrary processing or command-driving behavior.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The invocation guidance is overly broad, including general coding, testing, and deployment contexts rather than precise DNS/network troubleshooting triggers. In agent ecosystems, vague routing criteria can cause the skill to be selected in situations where its exec and network capabilities are unnecessary or unsafe.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The usage guidance lacks concrete trigger phrases and hard scope boundaries, so the agent may over-apply the skill to loosely related technical tasks. Because the skill advertises exec and network behavior, ambiguous selection logic increases the risk of unnecessary external connections or command execution.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill includes command examples that establish outbound connections and inspect remote certificates, but it does not prominently warn that using the skill may execute network diagnostic commands against external hosts. This is dangerous because users and higher-level agents may not realize the skill can trigger active network interactions, creating privacy, compliance, and unintended-target risks.

Static analysis

No suspicious patterns detected.