Multi Agent Protocol
Security checks across malware telemetry and agentic risk
Overview
This skill is a documented multi-agent workflow protocol, not executable code, and its state storage, approvals, and external-agent handoffs are disclosed and aligned with its purpose.
Before installing, make sure the task-store plugin enforces the permission matrix, keep human approval gates enabled for deployments, merges, external writes, and destructive edits, and avoid putting secrets or unnecessary private data into task specs, artifacts, or event logs.
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
66/66 vendors flagged this skill as clean.