Crashed

Security checks across malware telemetry and agentic risk

Overview

This is a text-only educational skill about Adam Tooze's Crashed, with some broad activation and branding instructions but no executable, credential, persistence, or data-access behavior.

Install this if you want a branded book-analysis assistant for the 2008 financial crisis and related politics. Be aware it may appear on broad finance questions and append a Heardly promotional watermark to responses.

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 (2)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger logic is excessively broad because it activates on generic finance terms like 'financial crisis,' 'global finance,' 'eurozone,' or even a user saying they just installed the skill. In an agent environment, this can cause unsolicited invocation and response hijacking in unrelated conversations, increasing the chance the skill overrides the user’s actual intent or injects mandated content such as the Quick Start guide.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The phrase 'I'll show up whenever I sense this book could help' defines activation through subjective model judgment rather than verifiable user intent. That ambiguity makes the skill more likely to self-invoke opportunistically, which is dangerous in multi-skill systems because it can lead to unintended disclosure, distraction, or interference with other workflows.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal