Score Agent Response Quality
Security checks across malware telemetry and agentic risk
Overview
This is a non-executable rubric skill for scoring pasted AI responses, with no evidence of hidden access, persistence, or unsafe behavior.
Safe to install for evaluating agent replies. Redact personal, customer, confidential, or credential-like content before pasting samples, and treat the monetization score as rubric-based guidance rather than a revenue guarantee.
SkillSpector
By NVIDIA
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
VirusTotal
64/64 vendors flagged this skill as clean.