iKuai 流量报表

• Outcome: Review
• Latest audit: May 29, 2026 at 9:39 PM UTC
• Version:
  1.0.2

Overview

This skill appears useful for router reporting, but it creates a broad, sensitive network report in a predictable temporary file without enough scoping or redaction.

Review before installing. Use this only for an iKuai/router environment you administer, treat the generated HTML as sensitive, avoid sharing it, delete it after use, and prefer a private output path with redaction for IPs, MAC addresses, hostnames, logs, DNS, VPN, and management-exposure details.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill documentation instructs users to run a Python script, invoke shell commands, and write an HTML file, but the skill does not declare corresponding permissions. This creates a transparency and policy gap: reviewers and users cannot accurately assess that the skill reads local auth/config state, executes external commands, and writes output, which increases the chance of unintended file or command side effects.

Description-Behavior Mismatch

Medium

• Confidence: 93% confidence
• Finding: This template goes well beyond a traffic visualization report and renders sensitive network/security configuration details such as WAN/LAN settings, DNS, VPN status, management exposure, and other operational metadata. In the context of an agent skill that may transform router JSON into shareable HTML, this creates an unnecessary information disclosure risk because the generated report can reveal attack-surface details useful for reconnaissance.

Description-Behavior Mismatch

Medium

• Confidence: 96% confidence
• Finding: The report embeds device inventory data, MAC addresses, hostnames, online client details, and system logs, all of which exceed the minimum data needed for a traffic report. If exported, shared, or exposed through the skill, these fields leak identifiable internal assets and activity history, enabling network mapping and targeted follow-on attacks.

Context-Inappropriate Capability

Low

• Confidence: 84% confidence
• Finding: Publishing the exact data collection commands and representative output schema is not necessary to display the final visualization and can disclose operational collection capabilities. While not directly exploitable by itself, it lowers the barrier for an observer to understand how to query the router, what telemetry exists, and what sensitive data may be retrievable.

Missing User Warnings

Low

• Confidence: 80% confidence
• Finding: The skill states a fixed output path of `/tmp/ikuai-report.html` and gives commands that generate the report there, but it does not warn that existing content at that path will be overwritten. While `/tmp` is a conventional temporary location, silent overwrite behavior can still cause data loss or confusion, especially on shared or automated systems.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The script writes a router report containing sensitive operational data such as IPs, logs, DNS, ACL/QoS counts, client details, and command-output snippets to a predictable file in /tmp. On multi-user systems or systems where /tmp is inspected, backed up, or shared through other tooling, this can expose internal network information without the user's explicit awareness.

VirusTotal

65/65 vendors flagged this skill as clean.