证券监管监控

• Outcome: Review
• Latest audit: May 29, 2026 at 9:33 PM UTC
• Version:
  1.0.0

Overview

The skill has a plausible public-page monitoring purpose, but it delegates core work to pre-existing root-level scripts and a persistent cron setup that are outside the reviewed package.

Install or run this only in an environment where you control and have reviewed the existing /root/monitoring/securities system. Before use, inspect the underlying crawler and notification scripts, confirm the cron entry is wanted, verify the XCrawl API key handling, check what is written to /tmp and /var/log, and confirm the Enterprise WeChat recipient before sending any notifications.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill explicitly instructs users to run scripts under /root, writes notification and diff artifacts into /tmp, and may send updates via enterprise messaging, but it does not clearly warn the user about privileged file writes, temporary-file handling, or external data transmission. This is dangerous because users may trigger actions with side effects on a privileged host or unintentionally exfiltrate monitored content to third-party messaging systems without informed consent.

VirusTotal

64/64 vendors flagged this skill as clean.