Security audit
Agent Reputation
Security checks across malware telemetry and agentic risk
Overview
This skill is a disclosed Agent Reputation integration that uses remote lookup and user-authorized write actions without hidden installation or persistence behavior.
Before installing, be aware that read-only discovery is unauthenticated, but registration, claiming, rating, feedback, and contact requests are external actions. Only provide an owner token when you intend to authorize those writes, and keep that token out of logs, URLs, source files, and chat transcripts.
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.
Static analysis
No suspicious patterns detected.