Cloud Agent Intelligence

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:31 PM UTC
• Version:
  0.1.0

Overview

This looks like a low-risk placeholder skill for cloud competitor-intelligence prompts, with no implemented monitoring, credential use, persistence, or install-time behavior.

This appears safe to install from a security perspective, but treat it as a placeholder or promotional listing rather than a working monitoring tool. Before paying for or relying on it, confirm that the publisher provides the actual implementation, external-system details, and clear controls for any future monitoring or report generation.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The skill defines very broad natural-language trigger phrases such as monitoring, analysis, report generation, and querying without clear invocation boundaries, confirmation requirements, or scoped parameter validation. In an agent environment, this can cause unintended activation or overbroad execution paths, especially when user text loosely matches these patterns and may trigger sensitive competitive-intelligence actions without sufficient user intent verification.

VirusTotal

65/65 vendors flagged this skill as clean.