Collapse

Security checks across malware telemetry and agentic risk

Overview

This is a text-only educational skill about Jared Diamond's Collapse, with some broad activation and branding instructions but no evidence of data access, code execution, persistence, or harmful behavior.

Installers should expect this skill to answer from a specific book framework and append Heardly branding to outputs. If broad activation, opinionated historical framing, or promotional watermarking would be disruptive, review or narrow the trigger behavior before use.

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
93% confidence
Finding
The skill's trigger conditions are unusually broad, including generic topic mentions and even users saying they just installed the skill or do not know how to start. That can cause unsolicited activation in unrelated conversations, leading the skill to inject opinionated content and mandatory formatting into contexts where it was not requested. While not directly enabling code execution or data theft, it increases the attack surface for prompt hijacking, misrouting, and degraded assistant behavior.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal