Security audit

DNS网络管理工具

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a DNS/network diagnostic helper, but it requests command execution while its scope and safety claims are broader than the artifact supports.

Review this skill before installing in sensitive environments. It appears non-malicious, but because it can run commands and its routing is broader than DNS/network diagnostics, only use it when you intentionally want live network checks and local diagnostic output.

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)

Intent-Code Divergence

Medium

Confidence: 92% confidence
Finding: The skill claims DNS-hijack protection via HTTPS-encrypted DNS requests, but the file only includes generic TLS certificate inspection commands and no actual DNS-over-HTTPS or DNS-over-TLS implementation. This creates a false sense of protection, which may cause users to trust insecure diagnostics or operational guidance under incorrect security assumptions.

Intent-Code Divergence

Medium

Confidence: 95% confidence
Finding: The manifest grants the skill `exec` capability, enabling arbitrary command execution, while the documentation advertises built-in safety/compliance and permission restrictions without showing any real guardrails. This mismatch is dangerous because users or orchestrators may over-trust the skill and allow execution in sensitive environments without the promised controls.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The activation guidance is overly broad and overlaps with general coding, debugging, deployment, and automation tasks unrelated to DNS/network diagnostics. Over-broad routing increases the chance this exec-capable skill is invoked in contexts where command execution is unnecessary, exposing systems to avoidable risk.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The examples and request schema are generic enough to fit many non-network workflows, which weakens invocation boundaries for a skill that can read files and execute commands. In agentic environments, ambiguous schemas can cause unintended selection and execution, increasing the attack surface and likelihood of unsafe actions.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The documentation describes command execution and writing output files such as certificate chains without clearly warning users about network access, filesystem effects, or the sensitivity of collected artifacts. In a skill with `exec` access, lack of clear warnings and constraints can lead to unintended system changes, data exposure, or misuse in production environments.

Static analysis

No suspicious patterns detected.