Agent Harness Construction
Security checks across malware telemetry and agentic risk
Overview
This skill is a short instructional guide for designing AI agent tools and does not install code or request sensitive access.
This appears reasonable to install as general agent-design guidance. Review any separate tools or workflows you later create from this advice, especially for deploys, migrations, permissions, or other high-impact actions.
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.