Pilot Digital Twin Setup

Security checks across malware telemetry and agentic risk

Overview

This is a setup guide for a four-agent digital twin deployment; its persistence and external integrations are disclosed enough to be treated as expected operational behavior.

Before installing, confirm that pilotctl and the pilot-* skills come from sources you trust, review any existing ~/.pilot setup before overwriting the manifest, and configure webhook or Slack endpoints only with appropriately scoped credentials.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
Findings (2)

Missing User Warnings

Low
Confidence
94% confidence
Finding
The setup procedure explicitly instructs writing a manifest under the user's home directory, but the skill description does not warn about this persistent filesystem change. This is a real transparency and consent issue because users may invoke the skill expecting transient setup guidance, not a lasting local configuration artifact.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The action-planner role is capable of sending maintenance orders and notifications to external systems, but this externally effectful behavior is not clearly disclosed in the top-level skill description. That omission is risky because users may unknowingly deploy a role that can trigger outbound business actions or notifications beyond the local agent environment.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal