Hyperion

Security checks across malware telemetry and agentic risk

Overview

This is a static Hyperion-themed guidance skill with some broad activation phrases, but it does not run code, access files, or request sensitive permissions.

Install this if you want Hyperion-based literary reflection and are comfortable with the assistant sometimes applying that frame to broad emotional prompts. Avoid relying on it for serious personal, medical, mental-health, legal, or financial decisions; treat it as thematic commentary, not professional advice.

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 (3)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The skill explicitly triggers when a user says they just installed the skill or does not know how to start, which is a broad onboarding/help condition unrelated to Hyperion-specific subject matter. This can cause the skill to activate in generic support or setup conversations, leading to unsolicited persona takeover and response hijacking in contexts where the user did not intend to invoke this book skill.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The invocation examples include very common emotional phrases like 'Facing the unknown,' 'My faith is shattered,' and 'A choice I made years ago still haunts me,' which are likely to appear in ordinary conversations far outside the intended book context. Because the skill is designed to proactively engage and steer responses, these broad triggers increase the chance of accidental activation during sensitive personal discussions, causing irrelevant or intrusive responses.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The activation guidance includes broad, emotionally common user statements such as losing faith, facing mystery, regret, caregiving decline, and feeling time pressure. These phrases can cause overbroad triggering in ordinary sensitive conversations, leading the assistant to steer users into a specific narrative frame when they may not have requested it, which is especially risky in emotionally charged contexts.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal