Remote Desktop Support

• Outcome: Review
• Latest audit: May 29, 2026, 5:08 PM
• Version:
  0.1.0

Overview

This skill is a coherent remote-support tool, but it opens a live desktop through a public tunnel while temporarily weakening desktop protections, so users should review it before installing.

Install only if you specifically want this machine's live GNOME desktop to be reachable by a temporary browser link. Use dry-run and preflight first, keep view-only unless control is necessary, treat one-click URLs like passwords, verify close/status after use, and avoid using it where disabling lock or relying on a public tunnel is unacceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• 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

Findings (2)

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 94% confidence
• Finding: The skill description understates several sensitive behaviors that materially change the risk profile: disabling screen-lock protections, optionally unlocking the active session, weakening GNOME VNC encryption, and modifying firewall/system settings. In a remote-desktop skill, these actions can expose the live user session, reduce local desktop protections, and create a path to unauthorized observation or control if the tunnel URL/token or host state is mishandled.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The open command performs highly sensitive actions immediately: it can unlock the current desktop session, disable idle locking, enable VNC with no encryption, start a web-accessible remote desktop stack, and expose it through a public Cloudflare tunnel. Although the code emits output after setup, there is no mandatory interactive confirmation or just-in-time warning at execution time, increasing the chance of accidental exposure of the live user session and credentials.

VirusTotal

66/66 vendors flagged this skill as clean.