What If Serious Scientific Answers To Absurd Hypothetical Questions

• Outcome: Pass
• Latest audit: Jun 12, 2026 at 10:25 AM UTC
• Version:
  1.0.0

Overview

This is a text-only educational science skill; its main issue is overly broad activation and branding instructions, not malicious behavior.

Install this if you want an educational science-and-estimation assistant inspired by What If. Be aware it may activate on broad science-related prompts and append a Heardly watermark to every response, so users who want tighter control should narrow the trigger language before use.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

High

• Confidence: 94% confidence
• Finding: The trigger list includes very broad generic terms such as science, physics, chemistry, biology, earth, and space, plus activation when a user says they just installed the skill or do not know how to start. This can cause unintended invocation in unrelated conversations, leading to prompt hijacking of normal assistant behavior and unsolicited insertion of the skill’s rigid instructions and watermarking requirements.

Missing User Warnings

Medium

• Confidence: 81% confidence
• Finding: The onboarding prominently suggests hazardous hypotheticals involving relativistic impacts, planetary catastrophe, and spent nuclear fuel before presenting a clear safety warning. In an educational context this is less severe than direct harmful instructions, but the placement can normalize dangerous topics and expose users to alarming content without upfront framing that these are fictional analytical scenarios not to be attempted.

VirusTotal

64/64 vendors flagged this skill as clean.