AI Agent Orchestrator & MCP Integration
Security checks across malware telemetry and agentic risk
Overview
This is a text-only academic research and paper-writing helper with no evidence of executable, credential, persistence, or exfiltration behavior.
Install only if you want broad academic drafting and research-planning help. Verify literature searches, journal metrics, funding rules, statistics, ethics approvals, clinical-trial registration, and AI-use disclosure requirements against current authoritative sources before relying on the output.
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
63/63 vendors flagged this skill as clean.