Security audit
Security checks for vulnerabilities and agentic risk
NetPing is a disclosed network diagnostic tool, but its scan commands should only be used on networks and hosts the user is authorized to test.
Install this only if you need shell-backed network diagnostics. Treat port checks, traceroute, HTTP probes, and especially subnet sweep as active network testing: use them only on systems you own or are authorized to assess, and expect the traffic to be visible to monitoring tools. I found no evidence of persistence, credential access, remote code loading, or destructive behavior.
The metadata describes a limited connectivity tool, but the documented commands expand into traceroute, HTTP probing, latency statistics, and subnet sweep host discovery. This mismatch can mislead reviewers and users about the real operational scope, causing the skill to be approved or invoked in contexts where broader reconnaissance behavior would not be acceptable.
The skill advertises shell-backed network operations but does not declare any tool scope, permissions, or allowed-tools constraints. That increases the risk of overbroad execution in agent environments, because a network-capable shell skill may run with more authority than intended and users or orchestrators are not clearly informed of its required privileges.
The skill includes port scanning, traceroute, HTTP checks, and subnet sweep functionality without any warning about authorization, privacy, rate limits, or possible impact on monitored networks. In practice, these features can trigger IDS/IPS alerts, violate policy, or be misused for reconnaissance against third-party systems if operators are not explicitly cautioned.
The tool performs active network operations such as pinging, port probing, traceroute, HTTP requests, DNS lookups, and host discovery without an explicit warning about their impact or the need for authorization. In an automated agent context, this can cause unintended external scanning or traffic generation, which may be inappropriate or policy-violating even if the commands are technically straightforward.
The script exposes traceroute, HTTP probing, and subnet sweep capabilities that are broader than the stated skill metadata and can generate outbound traffic beyond a simple ping, port check, or DNS lookup. This mismatch undermines informed consent and policy review, making it easier for a user or downstream agent to invoke scanning behavior that was not transparently declared.
The sweep command performs subnet-wide host discovery across 254 addresses, which is materially more aggressive than a single-host reachability test and can be repurposed for reconnaissance. In an agent skill, this expands the blast radius from diagnostic use into network enumeration, potentially violating least-privilege expectations or organizational scanning rules.
No suspicious patterns detected.
