V19 Agent Rating

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:44 PM UTC
• Version:
  1.0.0

Overview

This is a documentation-only skill for an agent-rating API, with user-run examples and no hidden code, but users should verify the external service before using its keys or registration endpoint.

Safe to install as documentation. Before running the curl commands, confirm you trust the Cloudflare-tunnel service, avoid sending sensitive agent names, and use only test or least-privilege governance keys you are comfortable sharing with that endpoint.

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: 98% confidence
• Finding: The skill explicitly publishes a usable API key and demonstrates how to call the remote service with it, without any warning about scope, rate limits, abuse, or revocation. Even if intended for public demo access, embedding a live credential in a distributable skill invites unauthorized use, quota exhaustion, monitoring blind spots, and downstream trust confusion if the key is reused or overprivileged.

VirusTotal

62/62 vendors flagged this skill as clean.