Adam Skill 1.0.0

Security checks across malware telemetry and agentic risk

Overview

This is a broad but disclosed personal-planning skill with no code, credentials, persistence, or hidden high-impact behavior.

Safe to install from the reviewed artifacts. Treat it as a general planning helper, and manually confirm any real-world actions involving messages, purchases, appointments, travel, or family information outside the skill.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (1)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill is defined as a broad, all-purpose personal assistant covering schedules, family matters, shopping, appointments, communications, planning, and decision support without clear boundaries or trigger constraints. This ambiguity can cause overbroad activation and make the agent handle sensitive personal contexts or high-impact tasks unexpectedly, increasing the chance of unsafe actions, privacy issues, or misuse.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal